conclusion of a concept paper

Concept Papers in Research: Deciphering the blueprint of brilliance

Concept papers hold significant importance as a precursor to a full-fledged research proposal in academia and research. Understanding the nuances and significance of a concept paper is essential for any researcher aiming to lay a strong foundation for their investigation.

Table of Contents

What Is Concept Paper

A concept paper can be defined as a concise document which outlines the fundamental aspects of a grant proposal. It outlines the initial ideas, objectives, and theoretical framework of a proposed research project. It is usually two to three-page long overview of the proposal. However, they differ from both research proposal and original research paper in lacking a detailed plan and methodology for a specific study as in research proposal provides and exclusion of the findings and analysis of a completed research project as in an original research paper. A concept paper primarily focuses on introducing the basic idea, intended research question, and the framework that will guide the research.

Purpose of a Concept Paper

A concept paper serves as an initial document, commonly required by private organizations before a formal proposal submission. It offers a preliminary overview of a project or research’s purpose, method, and implementation. It acts as a roadmap, providing clarity and coherence in research direction. Additionally, it also acts as a tool for receiving informal input. The paper is used for internal decision-making, seeking approval from the board, and securing commitment from partners. It promotes cohesive communication and serves as a professional and respectful tool in collaboration.

These papers aid in focusing on the core objectives, theoretical underpinnings, and potential methodology of the research, enabling researchers to gain initial feedback and refine their ideas before delving into detailed research.

Key Elements of a Concept Paper

Key elements of a concept paper include the title page , background , literature review , problem statement , methodology, timeline, and references. It’s crucial for researchers seeking grants as it helps evaluators assess the relevance and feasibility of the proposed research.

Writing an effective concept paper in academic research involves understanding and incorporating essential elements:

How to Write a Concept Paper?

To ensure an effective concept paper, it’s recommended to select a compelling research topic, pose numerous research questions and incorporate data and numbers to support the project’s rationale. The document must be concise (around five pages) after tailoring the content and following the formatting requirements. Additionally, infographics and scientific illustrations can enhance the document’s impact and engagement with the audience. The steps to write a concept paper are as follows:

1. Write a Crisp Title:

Choose a clear, descriptive title that encapsulates the main idea. The title should express the paper’s content. It should serve as a preview for the reader.

2. Provide a Background Information:

Give a background information about the issue or topic. Define the key terminologies or concepts. Review existing literature to identify the gaps your concept paper aims to fill.

3. Outline Contents in the Introduction:

Introduce the concept paper with a brief overview of the problem or idea you’re addressing. Explain its significance. Identify the specific knowledge gaps your research aims to address and mention any contradictory theories related to your research question.

4. Define a Mission Statement:

The mission statement follows a clear problem statement that defines the problem or concept that need to be addressed. Write a concise mission statement that engages your research purpose and explains why gaining the reader’s approval will benefit your field.

5. Explain the Research Aim and Objectives:

Explain why your research is important and the specific questions you aim to answer through your research. State the specific goals and objectives your concept intends to achieve. Provide a detailed explanation of your concept. What is it, how does it work, and what makes it unique?

6. Detail the Methodology:

Discuss the research methods you plan to use, such as surveys, experiments, case studies, interviews, and observations. Mention any ethical concerns related to your research.

7. Outline Proposed Methods and Potential Impact:

Provide detailed information on how you will conduct your research, including any specialized equipment or collaborations. Discuss the expected results or impacts of implementing the concept. Highlight the potential benefits, whether social, economic, or otherwise.

8. Mention the Feasibility

Discuss the resources necessary for the concept’s execution. Mention the expected duration of the research and specific milestones. Outline a proposed timeline for implementing the concept.

9. Include a Support Section:

Include a section that breaks down the project’s budget, explaining the overall cost and individual expenses to demonstrate how the allocated funds will be used.

10. Provide a Conclusion:

Summarize the key points and restate the importance of the concept. If necessary, include a call to action or next steps.

Although the structure and elements of a concept paper may vary depending on the specific requirements, you can tailor your document based on the guidelines or instructions you’ve been given.

Here are some tips to write a concept paper:

Example of a Concept Paper

Here is an example of a concept paper. Please note, this is a generalized example. Your concept paper should align with the specific requirements, guidelines, and objectives you aim to achieve in your proposal. Tailor it accordingly to the needs and context of the initiative you are proposing.

 Download Now!

Importance of a Concept Paper

Concept papers serve various fields, influencing the direction and potential of research in science, social sciences, technology, and more. They contribute to the formulation of groundbreaking studies and novel ideas that can impact societal, economic, and academic spheres.

A concept paper serves several crucial purposes in various fields:

In summary, a well-crafted concept paper is essential in outlining a clear, concise, and structured framework for new ideas or proposals. It helps in assessing the feasibility, viability, and potential impact of the concept before investing significant resources into its implementation.

How well do you understand concept papers? Test your understanding now! 

Fill the Details to Check Your Score

Role of AI in Writing Concept Papers

The increasing use of AI, particularly generative models, has facilitated the writing process for concept papers. Responsible use involves leveraging AI to assist in ideation, organization, and language refinement while ensuring that the originality and ethical standards of research are maintained.

AI plays a significant role in aiding the creation and development of concept papers in several ways:

1. Idea Generation and Organization

AI tools can assist in brainstorming initial ideas for concept papers based on key concepts. They can help in organizing information, creating outlines, and structuring the content effectively.

2. Summarizing Research and Data Analysis

AI-powered tools can assist in conducting comprehensive literature reviews, helping writers to gather and synthesize relevant information. AI algorithms can process and analyze vast amounts of data, providing insights and statistics to support the concept presented in the paper.

3. Language and Style Enhancement

AI grammar checker tools can help writers by offering grammar, style, and tone suggestions, ensuring professionalism. It can also facilitate translation, in case a global collaboration.

4. Collaboration and Feedback

AI platforms offer collaborative features that enable multiple authors to work simultaneously on a concept paper, allowing for real-time contributions and edits.

5. Customization and Personalization

AI algorithms can provide personalized recommendations based on the specific requirements or context of the concept paper. They can assist in tailoring the concept paper according to the target audience or specific guidelines.

6. Automation and Efficiency

AI can automate certain tasks, such as citation formatting, bibliography creation, or reference checking, saving time for the writer.

7. Analytics and Prediction

AI models can predict potential outcomes or impacts based on the information provided, helping writers anticipate the possible consequences of the proposed concept.

8. Real-Time Assistance

AI-driven chat-bots can provide real-time support and answers to specific questions related to the concept paper writing process.

AI’s role in writing concept papers significantly streamlines the writing process, enhances the quality of the content, and provides valuable assistance in various stages of development, contributing to the overall effectiveness of the final document.

Concept papers serve as the stepping stone in the research journey, aiding in the crystallization of ideas and the formulation of robust research proposals. It the cornerstone for translating ideas into impactful realities. Their significance spans diverse domains, from academia to business, enabling stakeholders to evaluate, invest, and realize the potential of groundbreaking concepts.

Frequently Asked Questions

A concept paper can be defined as a concise document outlining the fundamental aspects of a grant proposal such as the initial ideas, objectives, and theoretical framework of a proposed research project.

A good concept paper should offer a clear and comprehensive overview of the proposed research. It should demonstrate a strong understanding of the subject matter and outline a structured plan for its execution.

Concept paper is important to develop and clarify ideas, develop and evaluate proposal, inviting collaboration and collecting feedback, presenting proposals for academic and research initiatives and allocating resources.

I got wonderful idea

It helps a lot for my concept paper.

Rate this article Cancel Reply

Your email address will not be published.

Enago Academy's Most Popular Articles

• Reporting Research

Academic Essay Writing Made Simple: 4 types and tips

The pen is mightier than the sword, they say, and nowhere is this more evident…

Ensuring Academic Integrity and Transparency in Academic Research: A comprehensive checklist for researchers

Academic integrity is the foundation upon which the credibility and value of scientific findings are…

• AI in Academia

AI vs. AI: How to detect image manipulation and avoid academic misconduct

The scientific community is facing a new frontier of controversy as artificial intelligence (AI) is…

• Diversity and Inclusion

Need for Diversifying Academic Curricula: Embracing missing voices and marginalized perspectives

In classrooms worldwide, a single narrative often dominates, leaving many students feeling lost. These stories,…

• Career Corner
• Trending Now

Recognizing the signs: A guide to overcoming academic burnout

As the sun set over the campus, casting long shadows through the library windows, Alex…

Sign-up to read more

Subscribe for free to get unrestricted access to all our resources on research writing and academic publishing including:

• 2000+ blog articles
• 50+ Webinars
• 10+ Expert podcasts
• 50+ Infographics
• 10+ Checklists
• Research Guides

We hate spam too. We promise to protect your privacy and never spam you.

I am looking for Editing/ Proofreading services for my manuscript Tentative date of next journal submission:

As a researcher, what do you consider most when choosing an image manipulation detector?

Get science-backed answers as you write with Paperpal's Research feature

How to Write a Conclusion for Research Papers (with Examples)

The conclusion of a research paper is a crucial section that plays a significant role in the overall impact and effectiveness of your research paper. However, this is also the section that typically receives less attention compared to the introduction and the body of the paper. The conclusion serves to provide a concise summary of the key findings, their significance, their implications, and a sense of closure to the study. Discussing how can the findings be applied in real-world scenarios or inform policy, practice, or decision-making is especially valuable to practitioners and policymakers. The research paper conclusion also provides researchers with clear insights and valuable information for their own work, which they can then build on and contribute to the advancement of knowledge in the field.

The research paper conclusion should explain the significance of your findings within the broader context of your field. It restates how your results contribute to the existing body of knowledge and whether they confirm or challenge existing theories or hypotheses. Also, by identifying unanswered questions or areas requiring further investigation, your awareness of the broader research landscape can be demonstrated.

Remember to tailor the research paper conclusion to the specific needs and interests of your intended audience, which may include researchers, practitioners, policymakers, or a combination of these.

Table of Contents

What is a conclusion in a research paper, summarizing conclusion, editorial conclusion, externalizing conclusion, importance of a good research paper conclusion, how to write a conclusion for your research paper, research paper conclusion examples.

• How to write a research paper conclusion with Paperpal? 

Frequently Asked Questions

A conclusion in a research paper is the final section where you summarize and wrap up your research, presenting the key findings and insights derived from your study. The research paper conclusion is not the place to introduce new information or data that was not discussed in the main body of the paper. When working on how to conclude a research paper, remember to stick to summarizing and interpreting existing content. The research paper conclusion serves the following purposes: 1

• Warn readers of the possible consequences of not attending to the problem.
• Recommend specific course(s) of action.
• Restate key ideas to drive home the ultimate point of your research paper.
• Provide a “take-home” message that you want the readers to remember about your study.

Types of conclusions for research papers

In research papers, the conclusion provides closure to the reader. The type of research paper conclusion you choose depends on the nature of your study, your goals, and your target audience. I provide you with three common types of conclusions:

A summarizing conclusion is the most common type of conclusion in research papers. It involves summarizing the main points, reiterating the research question, and restating the significance of the findings. This common type of research paper conclusion is used across different disciplines.

An editorial conclusion is less common but can be used in research papers that are focused on proposing or advocating for a particular viewpoint or policy. It involves presenting a strong editorial or opinion based on the research findings and offering recommendations or calls to action.

An externalizing conclusion is a type of conclusion that extends the research beyond the scope of the paper by suggesting potential future research directions or discussing the broader implications of the findings. This type of conclusion is often used in more theoretical or exploratory research papers.

Align your conclusion’s tone with the rest of your research paper. Start Writing with Paperpal Now!  

The conclusion in a research paper serves several important purposes:

• Offers Implications and Recommendations : Your research paper conclusion is an excellent place to discuss the broader implications of your research and suggest potential areas for further study. It’s also an opportunity to offer practical recommendations based on your findings.
• Provides Closure : A good research paper conclusion provides a sense of closure to your paper. It should leave the reader with a feeling that they have reached the end of a well-structured and thought-provoking research project.
• Leaves a Lasting Impression : Writing a well-crafted research paper conclusion leaves a lasting impression on your readers. It’s your final opportunity to leave them with a new idea, a call to action, or a memorable quote.

Writing a strong conclusion for your research paper is essential to leave a lasting impression on your readers. Here’s a step-by-step process to help you create and know what to put in the conclusion of a research paper: 2

• Research Statement : Begin your research paper conclusion by restating your research statement. This reminds the reader of the main point you’ve been trying to prove throughout your paper. Keep it concise and clear.
• Key Points : Summarize the main arguments and key points you’ve made in your paper. Avoid introducing new information in the research paper conclusion. Instead, provide a concise overview of what you’ve discussed in the body of your paper.
• Address the Research Questions : If your research paper is based on specific research questions or hypotheses, briefly address whether you’ve answered them or achieved your research goals. Discuss the significance of your findings in this context.
• Significance : Highlight the importance of your research and its relevance in the broader context. Explain why your findings matter and how they contribute to the existing knowledge in your field.
• Implications : Explore the practical or theoretical implications of your research. How might your findings impact future research, policy, or real-world applications? Consider the “so what?” question.
• Future Research : Offer suggestions for future research in your area. What questions or aspects remain unanswered or warrant further investigation? This shows that your work opens the door for future exploration.
• Closing Thought : Conclude your research paper conclusion with a thought-provoking or memorable statement. This can leave a lasting impression on your readers and wrap up your paper effectively. Avoid introducing new information or arguments here.
• Proofread and Revise : Carefully proofread your conclusion for grammar, spelling, and clarity. Ensure that your ideas flow smoothly and that your conclusion is coherent and well-structured.

Write your research paper conclusion 2x faster with Paperpal. Try it now!

Remember that a well-crafted research paper conclusion is a reflection of the strength of your research and your ability to communicate its significance effectively. It should leave a lasting impression on your readers and tie together all the threads of your paper. Now you know how to start the conclusion of a research paper and what elements to include to make it impactful, let’s look at a research paper conclusion sample.

How to write a research paper conclusion with Paperpal?

A research paper conclusion is not just a summary of your study, but a synthesis of the key findings that ties the research together and places it in a broader context. A research paper conclusion should be concise, typically around one paragraph in length. However, some complex topics may require a longer conclusion to ensure the reader is left with a clear understanding of the study’s significance. Paperpal, an AI writing assistant trusted by over 800,000 academics globally, can help you write a well-structured conclusion for your research paper. 

• Sign Up or Log In: Create a new Paperpal account or login with your details.  
• Navigate to Features : Once logged in, head over to the features’ side navigation pane. Click on Templates and you’ll find a suite of generative AI features to help you write better, faster.  
• Generate an outline: Under Templates, select ‘Outlines’. Choose ‘Research article’ as your document type.  
• Select your section: Since you’re focusing on the conclusion, select this section when prompted.  
• Choose your field of study: Identifying your field of study allows Paperpal to provide more targeted suggestions, ensuring the relevance of your conclusion to your specific area of research. 
• Provide a brief description of your study: Enter details about your research topic and findings. This information helps Paperpal generate a tailored outline that aligns with your paper’s content. 
• Generate the conclusion outline: After entering all necessary details, click on ‘generate’. Paperpal will then create a structured outline for your conclusion, to help you start writing and build upon the outline.  
• Write your conclusion: Use the generated outline to build your conclusion. The outline serves as a guide, ensuring you cover all critical aspects of a strong conclusion, from summarizing key findings to highlighting the research’s implications. 
• Refine and enhance: Paperpal’s ‘Make Academic’ feature can be particularly useful in the final stages. Select any paragraph of your conclusion and use this feature to elevate the academic tone, ensuring your writing is aligned to the academic journal standards. 

By following these steps, Paperpal not only simplifies the process of writing a research paper conclusion but also ensures it is impactful, concise, and aligned with academic standards. Sign up with Paperpal today and write your research paper conclusion 2x faster .  

The research paper conclusion is a crucial part of your paper as it provides the final opportunity to leave a strong impression on your readers. In the research paper conclusion, summarize the main points of your research paper by restating your research statement, highlighting the most important findings, addressing the research questions or objectives, explaining the broader context of the study, discussing the significance of your findings, providing recommendations if applicable, and emphasizing the takeaway message. The main purpose of the conclusion is to remind the reader of the main point or argument of your paper and to provide a clear and concise summary of the key findings and their implications. All these elements should feature on your list of what to put in the conclusion of a research paper to create a strong final statement for your work.

A strong conclusion is a critical component of a research paper, as it provides an opportunity to wrap up your arguments, reiterate your main points, and leave a lasting impression on your readers. Here are the key elements of a strong research paper conclusion: 1. Conciseness : A research paper conclusion should be concise and to the point. It should not introduce new information or ideas that were not discussed in the body of the paper. 2. Summarization : The research paper conclusion should be comprehensive enough to give the reader a clear understanding of the research’s main contributions. 3 . Relevance : Ensure that the information included in the research paper conclusion is directly relevant to the research paper’s main topic and objectives; avoid unnecessary details. 4 . Connection to the Introduction : A well-structured research paper conclusion often revisits the key points made in the introduction and shows how the research has addressed the initial questions or objectives. 5. Emphasis : Highlight the significance and implications of your research. Why is your study important? What are the broader implications or applications of your findings? 6 . Call to Action : Include a call to action or a recommendation for future research or action based on your findings.

The length of a research paper conclusion can vary depending on several factors, including the overall length of the paper, the complexity of the research, and the specific journal requirements. While there is no strict rule for the length of a conclusion, but it’s generally advisable to keep it relatively short. A typical research paper conclusion might be around 5-10% of the paper’s total length. For example, if your paper is 10 pages long, the conclusion might be roughly half a page to one page in length.

In general, you do not need to include citations in the research paper conclusion. Citations are typically reserved for the body of the paper to support your arguments and provide evidence for your claims. However, there may be some exceptions to this rule: 1. If you are drawing a direct quote or paraphrasing a specific source in your research paper conclusion, you should include a citation to give proper credit to the original author. 2. If your conclusion refers to or discusses specific research, data, or sources that are crucial to the overall argument, citations can be included to reinforce your conclusion’s validity.

The conclusion of a research paper serves several important purposes: 1. Summarize the Key Points 2. Reinforce the Main Argument 3. Provide Closure 4. Offer Insights or Implications 5. Engage the Reader. 6. Reflect on Limitations

Remember that the primary purpose of the research paper conclusion is to leave a lasting impression on the reader, reinforcing the key points and providing closure to your research. It’s often the last part of the paper that the reader will see, so it should be strong and well-crafted.

• Makar, G., Foltz, C., Lendner, M., & Vaccaro, A. R. (2018). How to write effective discussion and conclusion sections. Clinical spine surgery, 31(8), 345-346.
• Bunton, D. (2005). The structure of PhD conclusion chapters.  Journal of English for academic purposes ,  4 (3), 207-224.

Paperpal is a comprehensive AI writing toolkit that helps students and researchers achieve 2x the writing in half the time. It leverages 21+ years of STM experience and insights from millions of research articles to provide in-depth academic writing, language editing, and submission readiness support to help you write better, faster.  

Get accurate academic translations, rewriting support, grammar checks, vocabulary suggestions, and generative AI assistance that delivers human precision at machine speed. Try for free or upgrade to Paperpal Prime starting at US$19 a month to access premium features, including consistency, plagiarism, and 30+ submission readiness checks to help you succeed.  

Experience the future of academic writing – Sign up to Paperpal and start writing for free!  

Related Reads:

• 5 Reasons for Rejection After Peer Review
• Ethical Research Practices For Research with Human Subjects

7 Ways to Improve Your Academic Writing Process

• Paraphrasing in Academic Writing: Answering Top Author Queries

Preflight For Editorial Desk: The Perfect Hybrid (AI + Human) Assistance Against Compromised Manuscripts

You may also like, how to write the first draft of a..., mla works cited page: format, template & examples, how to write a high-quality conference paper, academic editing: how to self-edit academic text with..., measuring academic success: definition & strategies for excellence, phd qualifying exam: tips for success , ai in education: it’s time to change the..., is it ethical to use ai-generated abstracts without..., what are journal guidelines on using generative ai..., quillbot review: features, pricing, and free alternatives.

• Walden University
• Faculty Portal

Writing a Paper: Conclusions

Writing a conclusion.

A conclusion is an important part of the paper; it provides closure for the reader while reminding the reader of the contents and importance of the paper. It accomplishes this by stepping back from the specifics in order to view the bigger picture of the document. In other words, it is reminding the reader of the main argument. For most course papers, it is usually one paragraph that simply and succinctly restates the main ideas and arguments, pulling everything together to help clarify the thesis of the paper. A conclusion does not introduce new ideas; instead, it should clarify the intent and importance of the paper. It can also suggest possible future research on the topic.

An Easy Checklist for Writing a Conclusion

It is important to remind the reader of the thesis of the paper so he is reminded of the argument and solutions you proposed.

Think of the main points as puzzle pieces, and the conclusion is where they all fit together to create a bigger picture. The reader should walk away with the bigger picture in mind.

Make sure that the paper places its findings in the context of real social change.

Make sure the reader has a distinct sense that the paper has come to an end. It is important to not leave the reader hanging. (You don’t want her to have flip-the-page syndrome, where the reader turns the page, expecting the paper to continue. The paper should naturally come to an end.)

No new ideas should be introduced in the conclusion. It is simply a review of the material that is already present in the paper. The only new idea would be the suggesting of a direction for future research.

Conclusion Example

As addressed in my analysis of recent research, the advantages of a later starting time for high school students significantly outweigh the disadvantages. A later starting time would allow teens more time to sleep--something that is important for their physical and mental health--and ultimately improve their academic performance and behavior. The added transportation costs that result from this change can be absorbed through energy savings. The beneficial effects on the students’ academic performance and behavior validate this decision, but its effect on student motivation is still unknown. I would encourage an in-depth look at the reactions of students to such a change. This sort of study would help determine the actual effects of a later start time on the time management and sleep habits of students.

Related Webinar

Didn't find what you need? Email us at [email protected] .

• Previous Page: Thesis Statements
• Next Page: Writer's Block
• Office of Student Disability Services

Walden Resources

Departments.

• Academic Residencies
• Academic Skills
• Career Planning and Development
• Customer Care Team
• Field Experience
• Military Services
• Student Success Advising
• Writing Skills

Centers and Offices

• Center for Social Change
• Office of Academic Support and Instructional Services
• Office of Degree Acceleration
• Office of Research and Doctoral Services
• Office of Student Affairs

Student Resources

• Doctoral Writing Assessment
• Form & Style Review
• Quick Answers
• ScholarWorks
• SKIL Courses and Workshops
• Walden Bookstore
• Walden Catalog & Student Handbook
• Student Safety/Title IX
• Legal & Consumer Information
• Website Terms and Conditions
• Cookie Policy
• Accessibility
• Accreditation
• State Authorization
• Net Price Calculator
• Contact Walden

Walden University is a member of Adtalem Global Education, Inc. www.adtalem.com Walden University is certified to operate by SCHEV © 2024 Walden University LLC. All rights reserved.

Community Blog

Keep up-to-date on postgraduate related issues with our quick reads written by students, postdocs, professors and industry leaders.

What is a Concept Paper and How do You Write One?

• By DiscoverPhDs
• August 26, 2020

What is a Concept Paper?

A concept paper is a short document written by a researcher before starting their research project, with the purpose of explaining what the study is about, why it is important and the methods that will be used.

The concept paper will include your proposed research title, a brief introduction to the subject, the aim of the study, the research questions you intend to answer, the type of data you will collect and how you will collect it. A concept paper can also be referred to as a research proposal.

What is the Purpose of a Concept Paper?

The primary aim of a research concept paper is to convince the reader that the proposed research project is worth doing. This means that the reader should first agree that the research study is novel and interesting. They should be convinced that there is a need for this research and that the research aims and questions are appropriate.

Finally, they should be satisfied that the methods for data collection proposed are feasible, are likely to work and can be performed within the specific time period allocated for this project.

The three main scenarios in which you may need to write a concept paper are if you are:

• A final year undergraduate or master’s student preparing to start a research project with a supervisor.
• A student submitting a research proposal to pursue a PhD project under the supervision of a professor.
• A principal investigator submitting a proposal to a funding body to secure financial support for a research project.

How Long is a Concept Paper?

The concept paper format is usually between 2 and 3 pages in length for students writing proposals for undergraduate, master’s or PhD projects. Concept papers written as part of funding applications may be over 20 pages in length.

How do you Write a Concept Paper?

There are 6 important aspects to consider when writing a concept paper or research proposal:

• 1. The wording of the title page, which is best presented as a question for this type of document. At this study concept stage, you can write the title a bit catchier, for example “Are 3D Printed Engine Parts Safe for Use in Aircraft?”.
• A brief introduction and review of relevant existing literature published within the subject area and identification of where the gaps in knowledge are. This last bit is particularly important as it guides you in defining the statement of the problem. The concept paper should provide a succinct summary of ‘the problem’, which is usually related to what is unknown or poorly understood about your research topic . By the end of the concept paper, the reader should be clear on how your research idea will provide a ‘solution’ to this problem.
• The overarching research aim of your proposed study and the objectives and/or questions you will address to achieve this aim. Align all of these with the problem statement; i.e. write each research question as a clear response to addressing the limitations and gaps identified from previous literature. Also give a clear description of your primary hypothesis.
• The specific data outputs that you plan to capture. For example, will this be qualitative or quantitative data? Do you plan to capture data at specific time points or at other defined intervals? Do you need to repeat data capture to asses any repeatability and reproducibility questions?
• The research methodology you will use to capture this data, including any specific measurement or analysis equipment and software you will use, and a consideration of statistical tests to help interpret the data. If your research requires the use of questionnaires, how will these be prepared and validated? In what sort of time frame would you plan to collect this data?
• Finally, include a statement of the significance of the study , explaining why your research is important and impactful. This can be in the form of a concluding paragraph that reiterate the statement of the problem, clarifies how your research will address this and explains who will benefit from your research and how.

You may need to include a short summary of the timeline for completing the research project. Defining milestones of the time points at which you intend to complete certain tasks can help to show that you’ve considered the practicalities of running this study. It also shows that what you have proposed is feasible in order to achieve your research goal.

If you’re pitching your proposed project to a funder, they may allocate a proportion of the money based on the satisfactory outcome of each milestone. These stakeholders may also be motivated by knowing that you intend to convert your dissertation into an article for journal publication; this level of dissemination is of high importance to them.

Additionally, you may be asked to provide a brief summary of the projected costs of running the study. For a PhD project this could be the bench fees associated with consumables and the cost of any travel if required.

Make sure to include references and cite all other literature and previous research that you discuss in your concept paper.

This guide gave you an overview of the key elements you need to know about when writing concept papers. The purpose of these are first to convey to the reader what your project’s purpose is and why your research topic is important; this is based on the development of a problem statement using evidence from your literature review.

Explain how it may positively impact your research field and if your proposed research design is appropriate and your planned research method achievable.

Starting your PhD can feel like a daunting, exciting and special time. They’ll be so much to think about – here are a few tips to help you get started.

Multistage sampling is a more complex form of cluster sampling for obtaining sample populations. Learn their pros and cons and how to undertake them.

The term monotonic relationship is a statistical definition that is used to describe the link between two variables.

Join thousands of other students and stay up to date with the latest PhD programmes, funding opportunities and advice.

Browse PhDs Now

Learn about defining your workspace, having a list of daily tasks and using technology to stay connected, all whilst working from home as a research student.

A well written figure legend will explain exactly what a figure means without having to refer to the main text. Our guide explains how to write one.

Alex is a PhD student at the University of Bradford researching ritual and funerary rites in later prehistoric Scotland: an analysis of faunal assemblages from the Covesea Caves.

Dr Ilesanmi has a PhD in Applied Biochemistry from the Federal University of Technology Akure, Ondo State, Nigeria. He is now a lecturer in the Department of Biochemistry at the Federal University Otuoke, Bayelsa State, Nigeria.

Join Thousands of Students

• Privacy Policy

Home » Research Paper Conclusion – Writing Guide and Examples

Research Paper Conclusion – Writing Guide and Examples

Table of Contents

Research Paper Conclusion

Definition:

A research paper conclusion is the final section of a research paper that summarizes the key findings, significance, and implications of the research. It is the writer’s opportunity to synthesize the information presented in the paper, draw conclusions, and make recommendations for future research or actions.

The conclusion should provide a clear and concise summary of the research paper, reiterating the research question or problem, the main results, and the significance of the findings. It should also discuss the limitations of the study and suggest areas for further research.

Parts of Research Paper Conclusion

The parts of a research paper conclusion typically include:

Restatement of the Thesis

The conclusion should begin by restating the thesis statement from the introduction in a different way. This helps to remind the reader of the main argument or purpose of the research.

Summary of Key Findings

The conclusion should summarize the main findings of the research, highlighting the most important results and conclusions. This section should be brief and to the point.

Implications and Significance

In this section, the researcher should explain the implications and significance of the research findings. This may include discussing the potential impact on the field or industry, highlighting new insights or knowledge gained, or pointing out areas for future research.

Limitations and Recommendations

It is important to acknowledge any limitations or weaknesses of the research and to make recommendations for how these could be addressed in future studies. This shows that the researcher is aware of the potential limitations of their work and is committed to improving the quality of research in their field.

Concluding Statement

The conclusion should end with a strong concluding statement that leaves a lasting impression on the reader. This could be a call to action, a recommendation for further research, or a final thought on the topic.

How to Write Research Paper Conclusion

Here are some steps you can follow to write an effective research paper conclusion:

• Restate the research problem or question: Begin by restating the research problem or question that you aimed to answer in your research. This will remind the reader of the purpose of your study.
• Summarize the main points: Summarize the key findings and results of your research. This can be done by highlighting the most important aspects of your research and the evidence that supports them.
• Discuss the implications: Discuss the implications of your findings for the research area and any potential applications of your research. You should also mention any limitations of your research that may affect the interpretation of your findings.
• Provide a conclusion : Provide a concise conclusion that summarizes the main points of your paper and emphasizes the significance of your research. This should be a strong and clear statement that leaves a lasting impression on the reader.
• Offer suggestions for future research: Lastly, offer suggestions for future research that could build on your findings and contribute to further advancements in the field.

Remember that the conclusion should be brief and to the point, while still effectively summarizing the key findings and implications of your research.

Example of Research Paper Conclusion

Here’s an example of a research paper conclusion:

Conclusion :

In conclusion, our study aimed to investigate the relationship between social media use and mental health among college students. Our findings suggest that there is a significant association between social media use and increased levels of anxiety and depression among college students. This highlights the need for increased awareness and education about the potential negative effects of social media use on mental health, particularly among college students.

Despite the limitations of our study, such as the small sample size and self-reported data, our findings have important implications for future research and practice. Future studies should aim to replicate our findings in larger, more diverse samples, and investigate the potential mechanisms underlying the association between social media use and mental health. In addition, interventions should be developed to promote healthy social media use among college students, such as mindfulness-based approaches and social media detox programs.

Overall, our study contributes to the growing body of research on the impact of social media on mental health, and highlights the importance of addressing this issue in the context of higher education. By raising awareness and promoting healthy social media use among college students, we can help to reduce the negative impact of social media on mental health and improve the well-being of young adults.

Purpose of Research Paper Conclusion

The purpose of a research paper conclusion is to provide a summary and synthesis of the key findings, significance, and implications of the research presented in the paper. The conclusion serves as the final opportunity for the writer to convey their message and leave a lasting impression on the reader.

The conclusion should restate the research problem or question, summarize the main results of the research, and explain their significance. It should also acknowledge the limitations of the study and suggest areas for future research or action.

Overall, the purpose of the conclusion is to provide a sense of closure to the research paper and to emphasize the importance of the research and its potential impact. It should leave the reader with a clear understanding of the main findings and why they matter. The conclusion serves as the writer’s opportunity to showcase their contribution to the field and to inspire further research and action.

When to Write Research Paper Conclusion

The conclusion of a research paper should be written after the body of the paper has been completed. It should not be written until the writer has thoroughly analyzed and interpreted their findings and has written a complete and cohesive discussion of the research.

Before writing the conclusion, the writer should review their research paper and consider the key points that they want to convey to the reader. They should also review the research question, hypotheses, and methodology to ensure that they have addressed all of the necessary components of the research.

Once the writer has a clear understanding of the main findings and their significance, they can begin writing the conclusion. The conclusion should be written in a clear and concise manner, and should reiterate the main points of the research while also providing insights and recommendations for future research or action.

Characteristics of Research Paper Conclusion

The characteristics of a research paper conclusion include:

• Clear and concise: The conclusion should be written in a clear and concise manner, summarizing the key findings and their significance.
• Comprehensive: The conclusion should address all of the main points of the research paper, including the research question or problem, the methodology, the main results, and their implications.
• Future-oriented : The conclusion should provide insights and recommendations for future research or action, based on the findings of the research.
• Impressive : The conclusion should leave a lasting impression on the reader, emphasizing the importance of the research and its potential impact.
• Objective : The conclusion should be based on the evidence presented in the research paper, and should avoid personal biases or opinions.
• Unique : The conclusion should be unique to the research paper and should not simply repeat information from the introduction or body of the paper.

Advantages of Research Paper Conclusion

The advantages of a research paper conclusion include:

• Summarizing the key findings : The conclusion provides a summary of the main findings of the research, making it easier for the reader to understand the key points of the study.
• Emphasizing the significance of the research: The conclusion emphasizes the importance of the research and its potential impact, making it more likely that readers will take the research seriously and consider its implications.
• Providing recommendations for future research or action : The conclusion suggests practical recommendations for future research or action, based on the findings of the study.
• Providing closure to the research paper : The conclusion provides a sense of closure to the research paper, tying together the different sections of the paper and leaving a lasting impression on the reader.
• Demonstrating the writer’s contribution to the field : The conclusion provides the writer with an opportunity to showcase their contribution to the field and to inspire further research and action.

Limitations of Research Paper Conclusion

While the conclusion of a research paper has many advantages, it also has some limitations that should be considered, including:

• I nability to address all aspects of the research: Due to the limited space available in the conclusion, it may not be possible to address all aspects of the research in detail.
• Subjectivity : While the conclusion should be objective, it may be influenced by the writer’s personal biases or opinions.
• Lack of new information: The conclusion should not introduce new information that has not been discussed in the body of the research paper.
• Lack of generalizability: The conclusions drawn from the research may not be applicable to other contexts or populations, limiting the generalizability of the study.
• Misinterpretation by the reader: The reader may misinterpret the conclusions drawn from the research, leading to a misunderstanding of the findings.

About the author

Muhammad Hassan

Researcher, Academic Writer, Web developer

You may also like

Limitations in Research – Types, Examples and...

Conceptual Framework – Types, Methodology and...

Problem Statement – Writing Guide, Examples and...

Research Paper Title Page – Example and Making...

Research Report – Example, Writing Guide and...

Delimitations in Research – Types, Examples and...

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

• PLOS Biology
• PLOS Climate
• PLOS Complex Systems
• PLOS Computational Biology
• PLOS Digital Health
• PLOS Genetics
• PLOS Global Public Health
• PLOS Medicine
• PLOS Mental Health
• PLOS Neglected Tropical Diseases
• PLOS Pathogens
• PLOS Sustainability and Transformation
• PLOS Collections
• How to Write Discussions and Conclusions

The discussion section contains the results and outcomes of a study. An effective discussion informs readers what can be learned from your experiment and provides context for the results.

What makes an effective discussion?

When you’re ready to write your discussion, you’ve already introduced the purpose of your study and provided an in-depth description of the methodology. The discussion informs readers about the larger implications of your study based on the results. Highlighting these implications while not overstating the findings can be challenging, especially when you’re submitting to a journal that selects articles based on novelty or potential impact. Regardless of what journal you are submitting to, the discussion section always serves the same purpose: concluding what your study results actually mean.

A successful discussion section puts your findings in context. It should include:

• the results of your research,
• a discussion of related research, and
• a comparison between your results and initial hypothesis.

Tip: Not all journals share the same naming conventions.

You can apply the advice in this article to the conclusion, results or discussion sections of your manuscript.

Our Early Career Researcher community tells us that the conclusion is often considered the most difficult aspect of a manuscript to write. To help, this guide provides questions to ask yourself, a basic structure to model your discussion off of and examples from published manuscripts. 

Questions to ask yourself:

• Was my hypothesis correct?
• If my hypothesis is partially correct or entirely different, what can be learned from the results? 
• How do the conclusions reshape or add onto the existing knowledge in the field? What does previous research say about the topic? 
• Why are the results important or relevant to your audience? Do they add further evidence to a scientific consensus or disprove prior studies? 
• How can future research build on these observations? What are the key experiments that must be done? 
• What is the “take-home” message you want your reader to leave with?

How to structure a discussion

Trying to fit a complete discussion into a single paragraph can add unnecessary stress to the writing process. If possible, you’ll want to give yourself two or three paragraphs to give the reader a comprehensive understanding of your study as a whole. Here’s one way to structure an effective discussion:

Writing Tips

While the above sections can help you brainstorm and structure your discussion, there are many common mistakes that writers revert to when having difficulties with their paper. Writing a discussion can be a delicate balance between summarizing your results, providing proper context for your research and avoiding introducing new information. Remember that your paper should be both confident and honest about the results! 

• Read the journal’s guidelines on the discussion and conclusion sections. If possible, learn about the guidelines before writing the discussion to ensure you’re writing to meet their expectations. 
• Begin with a clear statement of the principal findings. This will reinforce the main take-away for the reader and set up the rest of the discussion. 
• Explain why the outcomes of your study are important to the reader. Discuss the implications of your findings realistically based on previous literature, highlighting both the strengths and limitations of the research. 
• State whether the results prove or disprove your hypothesis. If your hypothesis was disproved, what might be the reasons? 
• Introduce new or expanded ways to think about the research question. Indicate what next steps can be taken to further pursue any unresolved questions. 
• If dealing with a contemporary or ongoing problem, such as climate change, discuss possible consequences if the problem is avoided. 
• Be concise. Adding unnecessary detail can distract from the main findings. 

Don’t

• Rewrite your abstract. Statements with “we investigated” or “we studied” generally do not belong in the discussion. 
• Include new arguments or evidence not previously discussed. Necessary information and evidence should be introduced in the main body of the paper. 
• Apologize. Even if your research contains significant limitations, don’t undermine your authority by including statements that doubt your methodology or execution. 
• Shy away from speaking on limitations or negative results. Including limitations and negative results will give readers a complete understanding of the presented research. Potential limitations include sources of potential bias, threats to internal or external validity, barriers to implementing an intervention and other issues inherent to the study design. 
• Overstate the importance of your findings. Making grand statements about how a study will fully resolve large questions can lead readers to doubt the success of the research. 

Snippets of Effective Discussions:

Consumer-based actions to reduce plastic pollution in rivers: A multi-criteria decision analysis approach

Identifying reliable indicators of fitness in polar bears

• How to Write a Great Title
• How to Write an Abstract
• How to Write Your Methods
• How to Report Statistics
• How to Edit Your Work

The contents of the Peer Review Center are also available as a live, interactive training session, complete with slides, talking points, and activities. …

The contents of the Writing Center are also available as a live, interactive training session, complete with slides, talking points, and activities. …

There’s a lot to consider when deciding where to submit your work. Learn how to choose a journal that will help your study reach its audience, while reflecting your values as a researcher…

• USC Libraries
• Research Guides

Organizing Your Social Sciences Research Paper

• 9. The Conclusion
• Purpose of Guide
• Design Flaws to Avoid
• Independent and Dependent Variables
• Glossary of Research Terms
• Reading Research Effectively
• Narrowing a Topic Idea
• Broadening a Topic Idea
• Extending the Timeliness of a Topic Idea
• Academic Writing Style
• Applying Critical Thinking
• Choosing a Title
• Making an Outline
• Paragraph Development
• Research Process Video Series
• Executive Summary
• The C.A.R.S. Model
• Background Information
• The Research Problem/Question
• Theoretical Framework
• Citation Tracking
• Content Alert Services
• Evaluating Sources
• Primary Sources
• Secondary Sources
• Tiertiary Sources
• Scholarly vs. Popular Publications
• Qualitative Methods
• Quantitative Methods
• Insiderness
• Using Non-Textual Elements
• Limitations of the Study
• Common Grammar Mistakes
• Writing Concisely
• Avoiding Plagiarism
• Footnotes or Endnotes?
• Further Readings
• Generative AI and Writing
• USC Libraries Tutorials and Other Guides
• Bibliography

The conclusion is intended to help the reader understand why your research should matter to them after they have finished reading the paper. A conclusion is not merely a summary of the main topics covered or a re-statement of your research problem, but a synthesis of key points derived from the findings of your study and, if applicable, where you recommend new areas for future research. For most college-level research papers, two or three well-developed paragraphs is sufficient for a conclusion, although in some cases, more paragraphs may be required in describing the key findings and their significance.

Conclusions. The Writing Center. University of North Carolina; Conclusions. The Writing Lab and The OWL. Purdue University.

Importance of a Good Conclusion

A well-written conclusion provides you with important opportunities to demonstrate to the reader your understanding of the research problem. These include:

• Presenting the last word on the issues you raised in your paper . Just as the introduction gives a first impression to your reader, the conclusion offers a chance to leave a lasting impression. Do this, for example, by highlighting key findings in your analysis that advance new understanding about the research problem, that are unusual or unexpected, or that have important implications applied to practice.
• Summarizing your thoughts and conveying the larger significance of your study . The conclusion is an opportunity to succinctly re-emphasize  your answer to the "So What?" question by placing the study within the context of how your research advances past research about the topic.
• Identifying how a gap in the literature has been addressed . The conclusion can be where you describe how a previously identified gap in the literature [first identified in your literature review section] has been addressed by your research and why this contribution is significant.
• Demonstrating the importance of your ideas . Don't be shy. The conclusion offers an opportunity to elaborate on the impact and significance of your findings. This is particularly important if your study approached examining the research problem from an unusual or innovative perspective.
• Introducing possible new or expanded ways of thinking about the research problem . This does not refer to introducing new information [which should be avoided], but to offer new insight and creative approaches for framing or contextualizing the research problem based on the results of your study.

Bunton, David. “The Structure of PhD Conclusion Chapters.” Journal of English for Academic Purposes 4 (July 2005): 207–224; Conclusions. The Writing Center. University of North Carolina; Kretchmer, Paul. Twelve Steps to Writing an Effective Conclusion. San Francisco Edit, 2003-2008; Conclusions. The Writing Lab and The OWL. Purdue University; Assan, Joseph. "Writing the Conclusion Chapter: The Good, the Bad and the Missing." Liverpool: Development Studies Association (2009): 1-8.

Structure and Writing Style

I.  General Rules

The general function of your paper's conclusion is to restate the main argument . It reminds the reader of the strengths of your main argument(s) and reiterates the most important evidence supporting those argument(s). Do this by clearly summarizing the context, background, and necessity of pursuing the research problem you investigated in relation to an issue, controversy, or a gap found in the literature. However, make sure that your conclusion is not simply a repetitive summary of the findings. This reduces the impact of the argument(s) you have developed in your paper.

When writing the conclusion to your paper, follow these general rules:

• Present your conclusions in clear, concise language. Re-state the purpose of your study, then describe how your findings differ or support those of other studies and why [i.e., what were the unique, new, or crucial contributions your study made to the overall research about your topic?].
• Do not simply reiterate your findings or the discussion of your results. Provide a synthesis of arguments presented in the paper to show how these converge to address the research problem and the overall objectives of your study.
• Indicate opportunities for future research if you haven't already done so in the discussion section of your paper. Highlighting the need for further research provides the reader with evidence that you have an in-depth awareness of the research problem but that further investigations should take place beyond the scope of your investigation.

Consider the following points to help ensure your conclusion is presented well:

• If the argument or purpose of your paper is complex, you may need to summarize the argument for your reader.
• If, prior to your conclusion, you have not yet explained the significance of your findings or if you are proceeding inductively, use the end of your paper to describe your main points and explain their significance.
• Move from a detailed to a general level of consideration that returns the topic to the context provided by the introduction or within a new context that emerges from the data [this is opposite of the introduction, which begins with general discussion of the context and ends with a detailed description of the research problem]. 

The conclusion also provides a place for you to persuasively and succinctly restate the research problem, given that the reader has now been presented with all the information about the topic . Depending on the discipline you are writing in, the concluding paragraph may contain your reflections on the evidence presented. However, the nature of being introspective about the research you have conducted will depend on the topic and whether your professor wants you to express your observations in this way. If asked to think introspectively about the topics, do not delve into idle speculation. Being introspective means looking within yourself as an author to try and understand an issue more deeply, not to guess at possible outcomes or make up scenarios not supported by the evidence.

II.  Developing a Compelling Conclusion

Although an effective conclusion needs to be clear and succinct, it does not need to be written passively or lack a compelling narrative. Strategies to help you move beyond merely summarizing the key points of your research paper may include any of the following:

• If your essay deals with a critical, contemporary problem, warn readers of the possible consequences of not attending to the problem proactively.
• Recommend a specific course or courses of action that, if adopted, could address a specific problem in practice or in the development of new knowledge leading to positive change.
• Cite a relevant quotation or expert opinion already noted in your paper in order to lend authority and support to the conclusion(s) you have reached [a good source would be from your literature review].
• Explain the consequences of your research in a way that elicits action or demonstrates urgency in seeking change.
• Restate a key statistic, fact, or visual image to emphasize the most important finding of your paper.
• If your discipline encourages personal reflection, illustrate your concluding point by drawing from your own life experiences.
• Return to an anecdote, an example, or a quotation that you presented in your introduction, but add further insight derived from the findings of your study; use your interpretation of results from your study to recast it in new or important ways.
• Provide a "take-home" message in the form of a succinct, declarative statement that you want the reader to remember about your study.

III. Problems to Avoid

Failure to be concise Your conclusion section should be concise and to the point. Conclusions that are too lengthy often have unnecessary information in them. The conclusion is not the place for details about your methodology or results. Although you should give a summary of what was learned from your research, this summary should be relatively brief, since the emphasis in the conclusion is on the implications, evaluations, insights, and other forms of analysis that you make. Strategies for writing concisely can be found here .

Failure to comment on larger, more significant issues In the introduction, your task was to move from the general [the field of study] to the specific [the research problem]. However, in the conclusion, your task is to move from a specific discussion [your research problem] back to a general discussion framed around the implications and significance of your findings [i.e., how your research contributes new understanding or fills an important gap in the literature]. In short, the conclusion is where you should place your research within a larger context [visualize your paper as an hourglass--start with a broad introduction and review of the literature, move to the specific analysis and discussion, conclude with a broad summary of the study's implications and significance].

Failure to reveal problems and negative results Negative aspects of the research process should never be ignored. These are problems, deficiencies, or challenges encountered during your study. They should be summarized as a way of qualifying your overall conclusions. If you encountered negative or unintended results [i.e., findings that are validated outside the research context in which they were generated], you must report them in the results section and discuss their implications in the discussion section of your paper. In the conclusion, use negative results as an opportunity to explain their possible significance and/or how they may form the basis for future research.

Failure to provide a clear summary of what was learned In order to be able to discuss how your research fits within your field of study [and possibly the world at large], you need to summarize briefly and succinctly how it contributes to new knowledge or a new understanding about the research problem. This element of your conclusion may be only a few sentences long.

Failure to match the objectives of your research Often research objectives in the social and behavioral sciences change while the research is being carried out. This is not a problem unless you forget to go back and refine the original objectives in your introduction. As these changes emerge they must be documented so that they accurately reflect what you were trying to accomplish in your research [not what you thought you might accomplish when you began].

Resist the urge to apologize If you've immersed yourself in studying the research problem, you presumably should know a good deal about it [perhaps even more than your professor!]. Nevertheless, by the time you have finished writing, you may be having some doubts about what you have produced. Repress those doubts! Don't undermine your authority as a researcher by saying something like, "This is just one approach to examining this problem; there may be other, much better approaches that...." The overall tone of your conclusion should convey confidence to the reader about the study's validity and realiability.

Assan, Joseph. "Writing the Conclusion Chapter: The Good, the Bad and the Missing." Liverpool: Development Studies Association (2009): 1-8; Concluding Paragraphs. College Writing Center at Meramec. St. Louis Community College; Conclusions. The Writing Center. University of North Carolina; Conclusions. The Writing Lab and The OWL. Purdue University; Freedman, Leora  and Jerry Plotnick. Introductions and Conclusions. The Lab Report. University College Writing Centre. University of Toronto; Leibensperger, Summer. Draft Your Conclusion. Academic Center, the University of Houston-Victoria, 2003; Make Your Last Words Count. The Writer’s Handbook. Writing Center. University of Wisconsin Madison; Miquel, Fuster-Marquez and Carmen Gregori-Signes. “Chapter Six: ‘Last but Not Least:’ Writing the Conclusion of Your Paper.” In Writing an Applied Linguistics Thesis or Dissertation: A Guide to Presenting Empirical Research . John Bitchener, editor. (Basingstoke,UK: Palgrave Macmillan, 2010), pp. 93-105; Tips for Writing a Good Conclusion. Writing@CSU. Colorado State University; Kretchmer, Paul. Twelve Steps to Writing an Effective Conclusion. San Francisco Edit, 2003-2008; Writing Conclusions. Writing Tutorial Services, Center for Innovative Teaching and Learning. Indiana University; Writing: Considering Structure and Organization. Institute for Writing Rhetoric. Dartmouth College.

Writing Tip

Don't Belabor the Obvious!

Avoid phrases like "in conclusion...," "in summary...," or "in closing...." These phrases can be useful, even welcome, in oral presentations. But readers can see by the tell-tale section heading and number of pages remaining that they are reaching the end of your paper. You'll irritate your readers if you belabor the obvious.

Assan, Joseph. "Writing the Conclusion Chapter: The Good, the Bad and the Missing." Liverpool: Development Studies Association (2009): 1-8.

Another Writing Tip

New Insight, Not New Information!

Don't surprise the reader with new information in your conclusion that was never referenced anywhere else in the paper. This why the conclusion rarely has citations to sources. If you have new information to present, add it to the discussion or other appropriate section of the paper. Note that, although no new information is introduced, the conclusion, along with the discussion section, is where you offer your most "original" contributions in the paper; the conclusion is where you describe the value of your research, demonstrate that you understand the material that you’ve presented, and position your findings within the larger context of scholarship on the topic, including describing how your research contributes new insights to that scholarship.

Assan, Joseph. "Writing the Conclusion Chapter: The Good, the Bad and the Missing." Liverpool: Development Studies Association (2009): 1-8; Conclusions. The Writing Center. University of North Carolina.

• << Previous: Limitations of the Study
• Next: Appendices >>
• Last Updated: May 30, 2024 9:38 AM
• URL: https://libguides.usc.edu/writingguide

So much is at stake in writing a conclusion. This is, after all, your last chance to persuade your readers to your point of view, to impress yourself upon them as a writer and thinker. And the impression you create in your conclusion will shape the impression that stays with your readers after they've finished the essay.

The end of an essay should therefore convey a sense of completeness and closure as well as a sense of the lingering possibilities of the topic, its larger meaning, its implications: the final paragraph should close the discussion without closing it off.

To establish a sense of closure, you might do one or more of the following:

• Conclude by linking the last paragraph to the first, perhaps by reiterating a word or phrase you used at the beginning.
• Conclude with a sentence composed mainly of one-syllable words. Simple language can help create an effect of understated drama.
• Conclude with a sentence that's compound or parallel in structure; such sentences can establish a sense of balance or order that may feel just right at the end of a complex discussion.

To close the discussion without closing it off, you might do one or more of the following:

• Conclude with a quotation from or reference to a primary or secondary source, one that amplifies your main point or puts it in a different perspective. A quotation from, say, the novel or poem you're writing about can add texture and specificity to your discussion; a critic or scholar can help confirm or complicate your final point. For example, you might conclude an essay on the idea of home in James Joyce's short story collection,  Dubliners , with information about Joyce's own complex feelings towards Dublin, his home. Or you might end with a biographer's statement about Joyce's attitude toward Dublin, which could illuminate his characters' responses to the city. Just be cautious, especially about using secondary material: make sure that you get the last word.
• Conclude by setting your discussion into a different, perhaps larger, context. For example, you might end an essay on nineteenth-century muckraking journalism by linking it to a current news magazine program like  60 Minutes .
• Conclude by redefining one of the key terms of your argument. For example, an essay on Marx's treatment of the conflict between wage labor and capital might begin with Marx's claim that the "capitalist economy is . . . a gigantic enterprise of dehumanization "; the essay might end by suggesting that Marxist analysis is itself dehumanizing because it construes everything in economic -- rather than moral or ethical-- terms.
• Conclude by considering the implications of your argument (or analysis or discussion). What does your argument imply, or involve, or suggest? For example, an essay on the novel  Ambiguous Adventure , by the Senegalese writer Cheikh Hamidou Kane, might open with the idea that the protagonist's development suggests Kane's belief in the need to integrate Western materialism and Sufi spirituality in modern Senegal. The conclusion might make the new but related point that the novel on the whole suggests that such an integration is (or isn't) possible.

Finally, some advice on how not to end an essay:

• Don't simply summarize your essay. A brief summary of your argument may be useful, especially if your essay is long--more than ten pages or so. But shorter essays tend not to require a restatement of your main ideas.
• Avoid phrases like "in conclusion," "to conclude," "in summary," and "to sum up." These phrases can be useful--even welcome--in oral presentations. But readers can see, by the tell-tale compression of the pages, when an essay is about to end. You'll irritate your audience if you belabor the obvious.
• Resist the urge to apologize. If you've immersed yourself in your subject, you now know a good deal more about it than you can possibly include in a five- or ten- or 20-page essay. As a result, by the time you've finished writing, you may be having some doubts about what you've produced. (And if you haven't immersed yourself in your subject, you may be feeling even more doubtful about your essay as you approach the conclusion.) Repress those doubts. Don't undercut your authority by saying things like, "this is just one approach to the subject; there may be other, better approaches. . ."

Copyright 1998, Pat Bellanca, for the Writing Center at Harvard University

How to Write a Conclusion for an Essay

By the time you get to the final paragraph of your paper, you have already done so much work on your essay, so all you want to do is to wrap it up as quickly as possible. You’ve already made a stunning introduction, proven your argument, and structured the whole piece as supposed – who cares about making a good conclusion paragraph?

The only thing you need to remember is that the conclusion of an essay is not just the last paragraph of an academic paper where you restate your thesis and key arguments. A concluding paragraph is also your opportunity to have a final impact on your audience. 

Feeling Overwhelmed Writing Your Essay Conclusion?

Simply send us your paper requirements, choose a writer and we’ll get it done fast.

How to write a conclusion paragraph that leaves a lasting impression – In this guide, the team at EssayPro is going to walk you through the process of writing a perfect conclusion step by step. Additionally, we will share valuable tips and tricks to help students of all ages impress their readers at the last moment.

Instead of Intro: What Is a Conclusion?

Before we can move on, let’s take a moment here to define the conclusion itself. According to the standard conclusion definition, it is pretty much the last part of something, its result, or end. However, this term is rather broad and superficial.

When it comes to writing academic papers, a concluding statement refers to an opinion, judgment, suggestion, or position arrived at by logical reasoning (through the arguments provided in the body of the text). Therefore, if you are wondering “what is a good closing sentence like?” – keep on reading.

What Does a Good Conclusion Mean?

Writing a good conclusion for a paper isn’t easy. However, we are going to walk you through this process step by step. Although there are generally no strict rules on how to formulate one, there are some basic principles that everyone should keep in mind. In this section, we will share some core ideas for writing a good conclusion, and, later in the article, we will also provide you with more practical advice and examples.

Here are the core goals a good conclusion should complete:

• “Wrap up” the entire paper;
• Demonstrate to readers that the author accomplished what he/she set out to do;
• Show how you the author has proved their thesis statement;
• Give a sense of completeness and closure on the topic;
• Leave something extra for your reader to think about;
• Leave a powerful final impact on a reader.

Another key thing to remember is that you should not introduce any new ideas or arguments to your paper's conclusion. It should only sum up what you have already written, revisit your thesis statement, and end with a powerful final impression.

When considering how to write a conclusion that works, here are the key points to keep in mind:

• A concluding sentence should only revisit the thesis statement, not restate it;
• It should summarize the main ideas from the body of the paper;
• It should demonstrate the significance and relevance of your work;
• An essay’s conclusion should include a call for action and leave space for further study or development of the topic (if necessary).

How Long Should a Conclusion Be? 

Although there are no strict universal rules regarding the length of an essay’s final clause, both teachers and experienced writers recommend keeping it clear, concise, and straight to the point. There is an unspoken rule that the introduction and conclusion of an academic paper should both be about 10% of the overall paper’s volume. For example, if you were assigned a 1500 word essay, both the introductory and final clauses should be approximately 150 words long (300 together).

Why You Need to Know How to End an Essay:

A conclusion is what drives a paper to its logical end. It also drives the main points of your piece one last time. It is your last opportunity to impact and impress your audience. And, most importantly, it is your chance to demonstrate to readers why your work matters. Simply put, the final paragraph of your essay should answer the last important question a reader will have – “So what?”

If you do a concluding paragraph right, it can give your readers a sense of logical completeness. On the other hand, if you do not make it powerful enough, it can leave them hanging, and diminish the effect of the entire piece.

Strategies to Crafting a Proper Conclusion

Although there are no strict rules for what style to use to write your conclusion, there are several strategies that have been proven to be effective. In the list below, you can find some of the most effective strategies with some good conclusion paragraph examples to help you grasp the idea.

One effective way to emphasize the significance of your essay and give the audience some thought to ponder about is by taking a look into the future. The “When and If” technique is quite powerful when it comes to supporting your points in the essay’s conclusion.

Prediction essay conclusion example: “Taking care of a pet is quite hard, which is the reason why most parents refuse their children’s requests to get a pet. However, the refusal should be the last choice of parents. If we want to inculcate a deep sense of responsibility and organization in our kids, and, at the same time, sprout compassion in them, we must let our children take care of pets.”

Another effective strategy is to link your conclusion to your introductory paragraph. This will create a full-circle narration for your readers, create a better understanding of your topic, and emphasize your key point.

Echo conclusion paragraph example: Introduction: “I believe that all children should grow up with a pet. I still remember the exact day my parents brought my first puppy to our house. This was one of the happiest moments in my life and, at the same time, one of the most life-changing ones. Growing up with a pet taught me a lot, and most importantly, it taught me to be responsible.” Conclusion:. “I remember when I picked up my first puppy and how happy I was at that time. Growing up with a pet, I learned what it means to take care of someone, make sure that he always has water and food, teach him, and constantly keep an eye on my little companion. Having a child grow up with a pet teaches them responsibility and helps them acquire a variety of other life skills like leadership, love, compassion, and empathy. This is why I believe that every kid should grow up with a pet!”

Finally, one more trick that will help you create a flawless conclusion is to amplify your main idea or to present it in another perspective of a larger context. This technique will help your readers to look at the problem discussed from a different angle.

Step-up argumentative essay conclusion example: “Despite the obvious advantages of owning a pet in childhood, I feel that we cannot generalize whether all children should have a pet. Whereas some kids may benefit from such experiences, namely, by becoming more compassionate, organized, and responsible, it really depends on the situation, motivation, and enthusiasm of a particular child for owning a pet.”

What is a clincher in an essay? – The final part of an essay’s conclusion is often referred to as a clincher sentence. According to the clincher definition, it is a final sentence that reinforces the main idea or leaves the audience with an intriguing thought to ponder upon. In a nutshell, the clincher is very similar to the hook you would use in an introductory paragraph. Its core mission is to seize the audience’s attention until the end of the paper. At the same time, this statement is what creates a sense of completeness and helps the author leave a lasting impression on the reader.

Now, since you now know what a clincher is, you are probably wondering how to use one in your own paper. First of all, keep in mind that a good clincher should be intriguing, memorable, smooth, and straightforward.

Generally, there are several different tricks you can use for your clincher statement; it can be:

• A short, but memorable and attention-grabbing conclusion;
• A relevant and memorable quote (only if it brings actual value);
• A call to action;
• A rhetorical question;
• An illustrative story or provocative example;
• A warning against a possibility or suggestion about the consequences of a discussed problem;
• A joke (however, be careful with this as it may not always be deemed appropriate).

Regardless of the technique you choose, make sure that your clincher is memorable and aligns with your introduction and thesis.

Clincher examples: - While New York may not be the only place with the breathtaking views, it is definitely among my personal to 3… and that’s what definitely makes it worth visiting. - “Thence we came forth to rebehold the stars”, Divine Comedy - Don’t you think all these advantages sound like almost life-saving benefits of owning a pet? “So we beat on, boats against the current, borne back ceaselessly into the past.”, The Great Gatsby

Conclusion Writing Don'ts 

Now, when you know what tricks and techniques you should use to create a perfect conclusion, let’s look at some of the things you should not do with our online paper writing service :

• Starting with some cliché concluding sentence starters. Many students find common phrases like “In conclusion,” “Therefore,” “In summary,” or similar statements to be pretty good conclusion starters. However, though such conclusion sentence starters may work in certain cases – for example, in speeches – they are overused, so it is recommended not to use them in writing to introduce your conclusion.
• Putting the first mention of your thesis statement in the conclusion – it has to be presented in your introduction first.
• Providing new arguments, subtopics, or ideas in the conclusion paragraph.
• Including a slightly changed or unchanged thesis statement.
• Providing arguments and evidence that belong in the body of the work.
• Writing too long, hard to read, or confusing sentences.

In case, you have written a conclusion, but you're not sure if it’s good enough?

EssayPro provides all kinds of writing assistance. Send your work to one of our top writers to get it reviewed in no time.

Conclusion Paragraph Outline

The total number of sentences in your final paragraph may vary depending on the number of points you discussed in your essay, as well as on the overall word count of your paper. However, the overall conclusion paragraph outline will remain the same and consists of the following elements:

• A conclusion starter:

The first part of your paragraph should drive readers back to your thesis statement. Thus, if you were wondering how to start a conclusion, the best way to do it is by rephrasing your thesis statement.

• Summary of the body paragraphs:

Right after revisiting your thesis, you should include several sentences that wrap up the key highlights and points from your body paragraphs. This part of your conclusion can consist of 2-3 sentences—depending on the number of arguments you’ve made. If necessary, you can also explain to the readers how your main points fit together.

• A concluding sentence:

Finally, you should end your paragraph with a last, powerful sentence that leaves a lasting impression, gives a sense of logical completeness, and connects readers back to the introduction of the paper.

These three key elements make up a perfect essay conclusion. Now, to give you an even better idea of how to create a perfect conclusion, let us give you a sample conclusion paragraph outline with examples from an argumentative essay on the topic of “Every Child Should Own a Pet:

• Sentence 1: Starter
• ~ Thesis: "Though taking care of a pet may be a bit challenging for small children. Parents should not restrict their kids from having a pet as it helps them grow into more responsible and compassionate people."
• ~ Restated thesis for a conclusion: "I can say that taking care of a pet is good for every child."
• Sentences 2-4: Summary
• ~ "Studies have shown that pet owners generally have fewer health problems."
• ~ "Owning a pet teaches a child to be more responsible."
• ~ "Spending time with a pet reduces stress, feelings of loneliness, and anxiety."
• Sentence 5: A concluding sentence
• ~ "Pets can really change a child life for the better, so don't hesitate to endorse your kid's desire to own a pet."

This is a clear example of how you can shape your conclusion paragraph.

How to Conclude Various Types of Essays

Depending on the type of academic essay you are working on, your concluding paragraph's style, tone, and length may vary. In this part of our guide, we will tell you how to end different types of essays and other works.

How to End an Argumentative Essay

Persuasive or argumentative essays always have the single goal of convincing readers of something (an idea, stance, or viewpoint) by appealing to arguments, facts, logic, and even emotions. The conclusion for such an essay has to be persuasive as well. A good trick you can use is to illustrate a real-life scenario that proves your stance or encourages readers to take action. More about persuasive essay outline you can read in our article.

Here are a few more tips for making a perfect conclusion for an argumentative essay:

• Carefully read the whole essay before you begin;
• Re-emphasize your ideas;
• Discuss possible implications;
• Don’t be afraid to appeal to the reader’s emotions.

How to End a Compare and Contrast Essay

The purpose of a compare and contrast essay is to emphasize the differences or similarities between two or more objects, people, phenomena, etc. Therefore, a logical conclusion should highlight how the reviewed objects are different or similar. Basically, in such a paper, your conclusion should recall all of the key common and distinctive features discussed in the body of your essay and also give readers some food for thought after they finish reading it.

How to Conclude a Descriptive Essay

The key idea of a descriptive essay is to showcase your creativity and writing skills by painting a vivid picture with the help of words. This is one of the most creative types of essays as it requires you to show a story, not tell it. This kind of essay implies using a lot of vivid details. Respectively, the conclusion of such a paper should also use descriptive imagery and, at the same time, sum up the main ideas. A good strategy for ending a descriptive essay would be to begin with a short explanation of why you wrote the essay. Then, you should reflect on how your topic affects you. In the middle of the conclusion, you should cover the most critical moments of the story to smoothly lead the reader into a logical closing statement. The “clincher”, in this case, should be a thought-provoking final sentence that leaves a good and lasting impression on the audience. Do not lead the reader into the essay and then leave them with dwindling memories of it.

How to Conclude an Essay About Yourself

If you find yourself writing an essay about yourself, you need to tell a personal story. As a rule, such essays talk about the author’s experiences, which is why a conclusion should create a feeling of narrative closure. A good strategy is to end your story with a logical finale and the lessons you have learned, while, at the same time, linking it to the introductory paragraph and recalling key moments from the story.

How to End an Informative Essay

Unlike other types of papers, informative or expository essays load readers with a lot of information and facts. In this case, “Synthesize, don’t summarize” is the best technique you can use to end your paper. Simply put, instead of recalling all of the major facts, you should approach your conclusion from the “So what?” position by highlighting the significance of the information provided.

How to Conclude a Narrative Essay

In a nutshell, a narrative essay is based on simple storytelling. The purpose of this paper is to share a particular story in detail. Therefore, the conclusion for such a paper should wrap up the story and avoid finishing on an abrupt cliffhanger. It is vital to include the key takeaways and the lessons learned from the story.

How to Write a Conclusion for a Lab Report

Unlike an essay, a lab report is based on an experiment. This type of paper describes the flow of a particular experiment conducted by a student and its conclusion should reflect on the outcomes of this experiment.

In thinking of how to write a conclusion for a lab, here are the key things you should do to get it right:

• Restate the goals of your experiment
• Describe the methods you used
• Include the results of the experiment and analyze the final data
• End your conclusion with a clear statement on whether or not the experiment was successful (Did you reach the expected results?)

How to Write a Conclusion for a Research Paper

Writing a paper is probably the hardest task of all, even for experienced dissertation writer . Unlike an essay or even a lab report, a research paper is a much longer piece of work that requires a deeper investigation of the problem. Therefore, a conclusion for such a paper should be even more sophisticated and powerful. If you're feeling difficulty writing an essay, you can buy essay on our service.

However, given that a research paper is the second most popular kind of academic paper (after an essay), it is important to know how to conclude a research paper. Even if you have not yet been assigned to do this task, be sure that you will face it soon. So, here are the steps you should follow to create a great conclusion for a research paper:

• Restate the Topic

Start your final paragraph with a quick reminder of what the topic of the piece is about. Keep it one sentence long.

• Revisit the Thesis

Next, you should remind your readers what your thesis statement was. However, do not just copy and paste it from the introductory clause: paraphrase your thesis so that you deliver the same idea but with different words. Keep your paraphrased thesis narrow, specific, and topic-oriented.

• Summarise Your Key Ideas

Just like the case of a regular essay’s conclusion, a research paper’s final paragraph should also include a short summary of all of the key points stated in the body sections. We recommend reading the entire body part a few times to define all of your main arguments and ideas.

• Showcase the Significance of Your Work

In the research paper conclusion, it is vital to highlight the significance of your research problem and state how your solution could be helpful.

• Make Suggestions for Future Studies

Finally, at the end of your conclusion, you should define how your findings will contribute to the development of its particular field of science. Outline the perspectives of further research and, if necessary, explain what is yet to be discovered on the topic.

Then, end your conclusion with a powerful concluding sentence – it can be a rhetorical question, call to action, or another hook that will help you have a strong impact on the audience.

• Answer the Right Questions

To create a top-notch research paper conclusion, be sure to answer the following questions:

• What is the goal of a research paper?
• What are the possible solutions to the research question(s)?
• How can your results be implemented in real life? (Is your research paper helpful to the community?)
• Why is this study important and relevant?

Additionally, here are a few more handy tips to follow:

• Provide clear examples from real life to help readers better understand the further implementation of the stated solutions;
• Keep your conclusion fresh, original, and creative.

Address to our term paper writers if you need to proofread or rewrite essay.

Want to Have Better Grades?

Address to our professionals and get your task done asap!

So, What Is a Good Closing Sentence? See The Difference

One of the best ways to learn how to write a good conclusion is to look at several professional essay conclusion examples. In this section of our guide, we are going to look at two different final paragraphs shaped on the basis of the same template, but even so, they are very different – where one is weak and the other is strong. Below, we are going to compare them to help you understand the difference between a good and a bad conclusion.

Here is the template we used: College degrees are in decline. The price of receiving an education does not correlate with the quality of the education received. As a result, graduated students face underemployment, and the worth of college degrees appears to be in serious doubt. However, the potential social and economic benefits of educated students balance out the equation.

Strong Conclusion ‍

People either see college as an opportunity or an inconvenience; therefore, a degree can only hold as much value as its owner’s skillset. The underemployment of graduate students puts the worth of college degrees in serious doubt. Yet, with the multitude of benefits that educated students bring to society and the economy, the equation remains in balance. Perhaps the ordinary person should consider college as a wise financial investment, but only if they stay determined to study and do the hard work.

Why is this example good? There are several key points that prove its effectiveness:

• There is a bold opening statement that encompasses the two contrasting types of students we can see today.
• There are two sentences that recall the thesis statement and cover the key arguments from the body of the essay.
• Finally, the last sentence sums up the key message of the essay and leaves readers with something to think about.

Weak Conclusion

In conclusion, with the poor preparation of students in college and the subsequent underemployment after graduation from college, the worth associated with the college degree appears to be in serious doubt. However, these issues alone may not reasonably conclude beyond a doubt that investing in a college degree is a rewarding venture. When the full benefits that come with education are carefully put into consideration and evaluated, college education for children in any country still has good advantages, and society should continue to advocate for a college education. The ordinary person should consider this a wise financial decision that holds rewards in the end. Apart from the monetary gains associated with a college education, society will greatly benefit from students when they finish college. Their minds are going to be expanded, and their reasoning and decision making will be enhanced.

What makes this example bad? Here are a few points to consider:

• Unlike the first example, this paragraph is long and not specific enough. The author provides plenty of generalized phrases that are not backed up by actual arguments.
• This piece is hard to read and understand and sentences have a confusing structure. Also, there are lots of repetitions and too many uses of the word “college”.
• There is no summary of the key benefits.
• The last two sentences that highlight the value of education contradict with the initial statement.
• Finally, the last sentence doesn’t offer a strong conclusion and gives no thought to ponder upon.
• In the body of your essay, you have hopefully already provided your reader(s) with plenty of information. Therefore, it is not wise to present new arguments or ideas in your conclusion.
• To end your final paragraph right, find a clear and straightforward message that will have the most powerful impact on your audience.
• Don’t use more than one quote in the final clause of your paper – the information from external sources (including quotes) belongs in the body of a paper.
• Be authoritative when writing a conclusion. You should sound confident and convincing to leave a good impression. Sentences like “I’m not an expert, but…” will most likely make you seem less knowledgeable and/or credible.

Good Conclusion Examples

Now that we've learned what a conclusion is and how to write one let's take a look at some essay conclusion examples to strengthen our knowledge.

The ending ironically reveals that all was for nothing. (A short explanation of the thematic effect of the book’s end) Tom says that Miss Watson freed Jim in her final will.Jim told Huck that the dead man on the Island was pap. The entire adventure seemingly evaporated into nothingness. (How this effect was manifested into the minds of thereaders).

All in all, international schools hold the key to building a full future that students can achieve. (Thesis statement simplified) They help students develop their own character by learning from their mistakes, without having to face a dreadful penalty for failure. (Thesis statement elaborated)Although some say that kids emerged “spoiled” with this mentality, the results prove the contrary. (Possible counter-arguments are noted)

In conclusion, public workers should be allowed to strike since it will give them a chance to air their grievances. (Thesis statement) Public workers should be allowed to strike when their rights, safety, and regulations are compromised. The workers will get motivated when they strike, and their demands are met.

In summary, studies reveal some similarities in the nutrient contents between the organic and non-organic food substances. (Starts with similarities) However, others have revealed many considerable differences in the amounts of antioxidants as well as other minerals present in organic and non-organic foods. Generally, organic foods have higher levels of antioxidants than non-organic foods and therefore are more important in the prevention of chronic illnesses.

As time went by, my obsession grew into something bigger than art; (‘As time went by’ signals maturation) it grew into a dream of developing myself for the world. (Showing student’s interest of developing himself for the community) It is a dream of not only seeing the world from a different perspective but also changing the perspective of people who see my work. (Showing student’s determination to create moving pieces of art)

In conclusion, it is evident that technology is an integral part of our lives and without it, we become “lost” since we have increasingly become dependent on its use. (Thesis with main point)

You might also be interested in reading nursing essay examples from our service.

How To Write A Conclusion For An Essay?

How to write a good conclusion, how to write a conclusion for a college essay.

Daniel Parker

is a seasoned educational writer focusing on scholarship guidance, research papers, and various forms of academic essays including reflective and narrative essays. His expertise also extends to detailed case studies. A scholar with a background in English Literature and Education, Daniel’s work on EssayPro blog aims to support students in achieving academic excellence and securing scholarships. His hobbies include reading classic literature and participating in academic forums.

is an expert in nursing and healthcare, with a strong background in history, law, and literature. Holding advanced degrees in nursing and public health, his analytical approach and comprehensive knowledge help students navigate complex topics. On EssayPro blog, Adam provides insightful articles on everything from historical analysis to the intricacies of healthcare policies. In his downtime, he enjoys historical documentaries and volunteering at local clinics.

Related Articles

• Departments and Units
• Majors and Minors
• LSA Course Guide
• LSA Gateway

Search: {{$root.lsaSearchQuery.q}}, Page {{$root.page}}

• Accessibility
• Undergraduates
• Instructors
• Alums & Friends

• ★ Writing Support
• Minor in Writing
• First-Year Writing Requirement
• Transfer Students
• Writing Guides
• Peer Writing Consultant Program
• Upper-Level Writing Requirement
• Writing Prizes
• International Students
• ★ The Writing Workshop
• Dissertation ECoach
• Fellows Seminar
• Dissertation Writing Groups
• Rackham / Sweetland Workshops
• Dissertation Writing Institute
• Guides to Teaching Writing
• Teaching Support and Services
• Support for FYWR Courses
• Support for ULWR Courses
• Writing Prize Nominating
• Alums Gallery
• Commencement
• Giving Opportunities
• How Do I Write an Intro, Conclusion, & Body Paragraph?
• How Do I Make Sure I Understand an Assignment?
• How Do I Decide What I Should Argue?
• How Can I Create Stronger Analysis?
• How Do I Effectively Integrate Textual Evidence?
• How Do I Write a Great Title?
• What Exactly is an Abstract?
• How Do I Present Findings From My Experiment in a Report?
• What is a Run-on Sentence & How Do I Fix It?
• How Do I Check the Structure of My Argument?
• How Do I Incorporate Quotes?
• How Can I Create a More Successful Powerpoint?
• How Can I Create a Strong Thesis?
• How Can I Write More Descriptively?
• How Do I Incorporate a Counterargument?
• How Do I Check My Citations?

See the bottom of the main Writing Guides page for licensing information.

Traditional Academic Essays In Three Parts

Part i: the introduction.

An introduction is usually the first paragraph of your academic essay. If you’re writing a long essay, you might need 2 or 3 paragraphs to introduce your topic to your reader. A good introduction does 2 things:

• Gets the reader’s attention. You can get a reader’s attention by telling a story, providing a statistic, pointing out something strange or interesting, providing and discussing an interesting quote, etc. Be interesting and find some original angle via which to engage others in your topic.
• Provides a specific and debatable thesis statement. The thesis statement is usually just one sentence long, but it might be longer—even a whole paragraph—if the essay you’re writing is long. A good thesis statement makes a debatable point, meaning a point someone might disagree with and argue against. It also serves as a roadmap for what you argue in your paper.

Part II: The Body Paragraphs

Body paragraphs help you prove your thesis and move you along a compelling trajectory from your introduction to your conclusion. If your thesis is a simple one, you might not need a lot of body paragraphs to prove it. If it’s more complicated, you’ll need more body paragraphs. An easy way to remember the parts of a body paragraph is to think of them as the MEAT of your essay:

Main Idea. The part of a topic sentence that states the main idea of the body paragraph. All of the sentences in the paragraph connect to it. Keep in mind that main ideas are…

• like labels. They appear in the first sentence of the paragraph and tell your reader what’s inside the paragraph.
• arguable. They’re not statements of fact; they’re debatable points that you prove with evidence.
• focused. Make a specific point in each paragraph and then prove that point.

Evidence. The parts of a paragraph that prove the main idea. You might include different types of evidence in different sentences. Keep in mind that different disciplines have different ideas about what counts as evidence and they adhere to different citation styles. Examples of evidence include…

• quotations and/or paraphrases from sources.
• facts , e.g. statistics or findings from studies you’ve conducted.
• narratives and/or descriptions , e.g. of your own experiences.

Analysis. The parts of a paragraph that explain the evidence. Make sure you tie the evidence you provide back to the paragraph’s main idea. In other words, discuss the evidence.

Transition. The part of a paragraph that helps you move fluidly from the last paragraph. Transitions appear in topic sentences along with main ideas, and they look both backward and forward in order to help you connect your ideas for your reader. Don’t end paragraphs with transitions; start with them.

Keep in mind that MEAT does not occur in that order. The “ T ransition” and the “ M ain Idea” often combine to form the first sentence—the topic sentence—and then paragraphs contain multiple sentences of evidence and analysis. For example, a paragraph might look like this: TM. E. E. A. E. E. A. A.

Part III: The Conclusion

A conclusion is the last paragraph of your essay, or, if you’re writing a really long essay, you might need 2 or 3 paragraphs to conclude. A conclusion typically does one of two things—or, of course, it can do both:

• Summarizes the argument. Some instructors expect you not to say anything new in your conclusion. They just want you to restate your main points. Especially if you’ve made a long and complicated argument, it’s useful to restate your main points for your reader by the time you’ve gotten to your conclusion. If you opt to do so, keep in mind that you should use different language than you used in your introduction and your body paragraphs. The introduction and conclusion shouldn’t be the same.
• For example, your argument might be significant to studies of a certain time period .
• Alternately, it might be significant to a certain geographical region .
• Alternately still, it might influence how your readers think about the future . You might even opt to speculate about the future and/or call your readers to action in your conclusion.

Handout by Dr. Liliana Naydan. Do not reproduce without permission.

• Information For
• Prospective Students
• Current Students
• Faculty and Staff
• Alumni and Friends
• More about LSA
• How Do I Apply?
• LSA Magazine
• Student Resources
• Academic Advising
• Global Studies
• LSA Opportunity Hub
• Social Media
• Update Contact Info
• Privacy Statement
• Report Feedback

• PRO Courses Guides New Tech Help Pro Expert Videos About wikiHow Pro Upgrade Sign In
• EDIT Edit this Article
• EXPLORE Tech Help Pro About Us Random Article Quizzes Request a New Article Community Dashboard This Or That Game Popular Categories Arts and Entertainment Artwork Books Movies Computers and Electronics Computers Phone Skills Technology Hacks Health Men's Health Mental Health Women's Health Relationships Dating Love Relationship Issues Hobbies and Crafts Crafts Drawing Games Education & Communication Communication Skills Personal Development Studying Personal Care and Style Fashion Hair Care Personal Hygiene Youth Personal Care School Stuff Dating All Categories Arts and Entertainment Finance and Business Home and Garden Relationship Quizzes Cars & Other Vehicles Food and Entertaining Personal Care and Style Sports and Fitness Computers and Electronics Health Pets and Animals Travel Education & Communication Hobbies and Crafts Philosophy and Religion Work World Family Life Holidays and Traditions Relationships Youth
• Browse Articles
• Learn Something New
• Quizzes Hot
• This Or That Game
• Train Your Brain
• Explore More
• Support wikiHow
• About wikiHow
• Log in / Sign up
• Finance and Business
• Business Skills
• Business Writing

How to Write a Concept Paper

Last Updated: March 20, 2023 Fact Checked

This article was co-authored by wikiHow Staff . Our trained team of editors and researchers validate articles for accuracy and comprehensiveness. wikiHow's Content Management Team carefully monitors the work from our editorial staff to ensure that each article is backed by trusted research and meets our high quality standards. This article has been fact-checked, ensuring the accuracy of any cited facts and confirming the authority of its sources. This article has been viewed 1,628,978 times. Learn more...

If you’ve got a great idea for a new product, program, or service, writing a concept paper is one way to seek funding for it. Concept papers describe the purpose and projected outcomes of the project, and are delivered to potential sponsors. To create a successful one, use clear, passionate language that expresses why your project matters, and who will benefit from it. Above all, show the sponsor that the goals of your project match up with the kinds of initiatives they want to support.

Sample Concept Papers

Establishing the Purpose

• For instance, you could start off your paper with an attention-grabbing statistic related to your project: “Every year, 10.5 million pounds of food go to waste due to one common pest: rats.”
• Giving your concept paper a descriptive title, like “Lock the Rat Box: Humane, Hands-Free Rodent Control,” is another good way to grab their attention.

• Try something like: “The Savco Foundation has long been committed to funding projects that foster healthy communities. We have developed Lock the Rat Box as an easy, cost-effective means to lower illness rates and sanitation costs in municipalities, and are seeking your support for the project.”

• For instance, your concept paper could include a statement like: “Rats are a nuisance, but also a serious vector of diseases such as rabies and the bubonic plague. Municipalities across the United States spend upwards of twenty million dollars a year combating these issues.”
• Include references to verify any data you cite.

Explaining How your Concept Works

• For instance, your project may involve building a prototype device to humanely trap rats.
• Your methods might also involve activities. For instance, you may propose advertising programs to educate communities about rat problems, or sending investigators to study the extent of the issue in various communities.

• Try using statements like: “While previous governmental services have explained rat infestations via poster, radio, and television campaigns, they have not taken advantage of social media as a means of connecting with community members. Our project fills that gap.”

• For example: “February 2018: sign a lease for a workshop space. Late February 2018: purchase materials for Lock the Rat Box prototype. March 2018: conduct preliminary tests of the prototype.”

• Other assessment tools could include things like surveys to gauge customer satisfaction, community involvement, or other metrics.

• Personnel, including any assistants
• Equipment and supplies
• Consultants you may need to bring in
• Space (rent, for example)

Reviewing the Draft

• If the application requests a particular format, follow the directions exactly.
• Otherwise, type your paper in a standard font at a readable size (12 point is good), number your pages, and use reasonable margins (1 inch all around is fine).

• For instance, avoid statements like “We believe that our product, Lock the Rat Box, could potentially help certain municipalities at least control rat infestations.”
• A stronger statement would be: “Lock the Rat Box will curtail rat infestations in any mid-sized municipality, and completely eradicate them in many cases.”

• If you are writing for a general, non-expert audience, ask someone unfamiliar with your project to read your concept paper and tell you if there were any parts they did not understand.

• Have someone who has not previously read your concept paper take a look at the final draft before you submit it. They’ll be more likely to catch any lingering errors.

Community Q&A

You Might Also Like

• ↑ https://www.aub.edu.lb/ogc/Documents/Writing_Concept_Paper.pdf
• ↑ https://ovpr.uconn.edu/wp-content/uploads/sites/2557/2018/09/How-to-Write-a-Concept-Paper.pdf
• ↑ https://www.ias.edu/sites/default/files/media-assets/Guidance%20Doc_Concept%20Paper.pdf
• ↑ https://www.umass.edu/cfr/grant-writing/guidelines-letter-intent

About This Article

To write a concept paper for a new product, program, or service, start with a descriptive, attention-grabbing title. Then, explain why you’re approaching the sponsor by describing what your project goals have in common with their company mission. Next, describe the problem you want to solve, and the methods you’ll use in order to solve it. Additionally, include a timeline for implementing your methods, and a preliminary budget with a list of the estimated costs. Finally, end your paper with a short summary reiterating your project’s purpose. For more advice, including how to make your paper stand out with proper formatting and action-oriented language, keep reading. Did this summary help you? Yes No

• Send fan mail to authors

Reader Success Stories

Kathryn Torres

Jul 17, 2020

Did this article help you?

Ongole Churchill

Jun 17, 2016

Ruth Denson

Mar 6, 2020

Naseem Khan

Apr 4, 2019

Kumudham Sandrasegaran

Oct 5, 2018

Featured Articles

Trending Articles

Watch Articles

• Terms of Use
• Privacy Policy
• Do Not Sell or Share My Info
• Not Selling Info

wikiHow Tech Help Pro:

Develop the tech skills you need for work and life

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 27 May 2024

Research on domain ontology construction based on the content features of online rumors

• Jianbo Zhao 1 ,
• Huailiang Liu 1 ,
• Weili Zhang 1 ,
• Tong Sun 1 ,
• Qiuyi Chen 1 ,
• Yuehai Wang 2 ,
• Jiale Cheng 2 ,
• Yan Zhuang 1 ,
• Xiaojin Zhang 1 ,
• Shanzhuang Zhang 1 ,
• Bowei Li 3 &
• Ruiyu Ding 2  

Scientific Reports volume  14 , Article number:  12134 ( 2024 ) Cite this article

229 Accesses

1 Altmetric

Metrics details

• Computational neuroscience
• Computer science
• Data acquisition
• Data integration
• Data mining
• Data processing
• Human behaviour
• Information technology
• Literature mining
• Machine learning
• Scientific data

Online rumors are widespread and difficult to identify, which bring serious harm to society and individuals. To effectively detect and govern online rumors, it is necessary to conduct in-depth semantic analysis and understand the content features of rumors. This paper proposes a TFI domain ontology construction method, which aims to achieve semantic parsing and reasoning of the rumor text content. This paper starts from the term layer, the frame layer, and the instance layer, and based on the reuse of the top-level ontology, the extraction of core literature content features, and the discovery of new concepts in the real corpus, obtains the core classes (five parent classes and 88 subclasses) of the rumor domain ontology and defines their concept hierarchy. Object properties and data properties are designed to describe relationships between entities or their features, and the instance layer is created according to the real rumor datasets. OWL language is used to encode the ontology, Protégé is used to visualize it, and SWRL rules and pellet reasoner are used to mine and verify implicit knowledge of the ontology, and judge the category of rumor text. This paper constructs a rumor domain ontology with high consistency and reliability.

Similar content being viewed by others

From rumor to genetic mutation detection with explanations: a GAN approach

Improving long COVID-related text classification: a novel end-to-end domain-adaptive paraphrasing framework

How do we study misogyny in the digital age? A systematic literature review using a computational linguistic approach

Introduction.

Online rumors are false information spread through online media, which have the characteristics of wide content 1 , hard to identify 2 , 3 . Online rumors can mislead the public, disrupt social order, damage personal and collective reputations, and pose a great challenge to the governance of internet information content. Therefore, in order to effectively detect and govern online rumors, it is necessary to conduct an in-depth semantic analysis and understanding of the rumor text content features.

The research on the content features of online rumors focuses on the lexical, syntactic and semantic features of the rumor text, including lexical, syntactic and semantic features 4 , syntactic structure and functional features 5 , source features 5 , 6 , rhetorical methods 7 , narrative structure 6 , 7 , 8 , language style 6 , 9 , 10 , corroborative means 10 , 11 and emotional features 10 , 12 , 13 , 14 , 15 , 16 , 17 , 18 . Most of the existing researches on rumor content features are feature mining under a single domain topic type, and lack of mining the influence relationship between multiple features. Therefore, this paper proposes to build an online rumor domain ontology to realize fine-grained hierarchical modeling of the relationship between rumor content features and credible verification of its effectiveness. Domain ontology is a systematic description of the objective existence in a specific discipline 19 . The construction methods mainly include TOVE method 20 , skeleton method 21 , IDEF-5 method 22 , 23 , methontology method 24 , 25 and seven-step method 26 , 27 , among which seven-step method is the most mature and widely used method at present 28 , which has strong systematicness and applicability 29 , but it does not provide quantitative indicators and methods about the quality and effect of ontology. The construction technology can be divided into the construction technology based on thesaurus conversion, the construction technology based on existing ontology reuse and the semi-automatic and automatic construction technology based on ontology engineering method 30 . The construction technology based on thesaurus conversion and the construction technology based on existing ontology reuse can save construction time and cost, and improve ontology reusability and interoperability, but there are often differences in structure, semantics and scene. Semi-automatic and automatic construction technology based on ontology engineering method The application of artificial intelligence technology can automatically extract ontology elements and structures from data sources with high efficiency and low cost, but the quality and accuracy are difficult to guarantee. Traditional domain ontology construction methods lack effective quality evaluation support, and construction technology lacks effective integration application. Therefore, this paper proposes an improved TFI network rumor domain ontology construction method based on the seven-step method. Starting from the terminology layer, the framework layer and the instance layer, it integrates the top-level ontology and core document content feature reuse technology, the bottom-up semi-automatic construction technology based on N-gram new word discovery algorithm and RoBERTa-Kmeans clustering algorithm, defines the fine-grained features of network rumor content and carries out hierarchical modeling. Using SWRL rules and pellet inference machine, the tacit knowledge of ontology is mined, and the quality of ontology validity and consistency is evaluated and verified.

The structure of this paper is as follows: Sect “ Related work ” introduces the characteristics of rumor content and the related work of domain ontology construction.; Sect “ Research method ” constructs the term layer, the frame layer and the instance layer of the domain ontology; Sect “ Domain ontology construction ” mines and verifies the implicit knowledge of the ontology based on SWRL rules and Pellet reasoner; Sect “ Ontology reasoning and validation ” points out the research limitations and future research directions; Sect “ Discussion ” summarizes the research content and contribution; Sect “ Conclusion ” summarizes the research content and contribution of this paper.

Related Work

Content features of online rumors.

The content features of online rumors refer to the adaptive description of vocabulary, syntax and semantics in rumor texts. Fu et al. 5 have made a linguistic analysis of COVID-19’s online rumors from the perspectives of pragmatics, discourse analysis and syntax, and concluded that the source of information, the specific place and time of the event, the length of the title and statement, and the emotions aroused are the important characteristics to judge the authenticity of the rumors; Zhang et al. 6 summarized the narrative theme, narrative characteristics, topic characteristics, language style and source characteristics of new media rumors; Li et al. 7 found that rumors have authoritative blessing and fear appeal in headline rhetoric, and they use news and digital headlines extensively, and the topic construction mostly uses programmed fixed structure; Yu et al. 8 analyzed and summarized the content distribution, narrative structure, topic scene construction and title characteristics of rumors in detail; Mourao et al. 9 found that the language style of rumors is significantly different from that of real texts, and rumors tend to use simpler, more emotional and more radical discourse strategies; Zhou et al. 10 analyzed the rumor text based on six analysis categories, such as content type, focus object and corroboration means, and found that the epidemic rumors were mostly “infectious” topics, with narrative expression being the most common, strong fear, and preference for exaggerated and polarized discourse style. Huang et al. 11 conducted an empirical study based on WeChat rumors, and found that the “confirmation” means of rumors include data corroboration and specific information, hot events and authoritative release; Butt et al. 12 analyzed the psycholinguistic features of rumors, and extracted four features from the rumor data set: LIWC, readability, senticnet and emotions. Zhou et al. 13 analyzed the semantic features of fake news content in theme and emotion, and found that the distribution of fake news and real news is different in theme features, and the overall mood, negative mood and anger of fake news are higher; Tan et al. 14 divided the content characteristics of rumors into content characteristics with certain emotional tendency and social characteristics that affect credibility; Damstra et al. 15 identified the elements as a consistent indicator of intentionally deceptive news content, including negative emotions causing anger or fear, lengthy sensational headlines, using informal language or swearing, etc. Lai et al. 16 put forward that emotional rumors can make the rumor audience have similar positive and negative emotions through emotional contagion; Yuan et al. 17 found that multimedia evidence form and topic shaping are important means to create rumors, which mostly convey negative emotions of fear and anger, and the provision of information sources is related to the popularity and duration of rumors; Ruan et al. 18 analyzed the content types, emotional types and discourse focus of Weibo’s rumor samples, and found that the proportion of social life rumors was the highest, and the emotional types were mainly hostile and fearful, with the focus on the general public and the personnel of the party, government and military institutions.

The forms and contents of online rumors tend to be diversified and complicated. The existing research on the content features of rumors is mostly aimed at the mining of content characteristics under specific topics, which cannot cover various types of rumor topics, and lacks fine-grained hierarchical modeling of the relationship between features and credible verification of their effectiveness.

Domain ontology construction

Domain ontology is a unified definition, standardized organization and visual representation of the concepts of knowledge in a specific domain 31 , 32 , and it is an important source of information for knowledge-based systems 19 , 33 . Theoretical methods include TOVE method 20 , skeleton method 21 , IDEF-5 method 22 , 23 , methontology method 24 , 25 and seven-step method 26 , 27 . TOVE method transforms informal description into formal ontology, which is suitable for fields that need accurate knowledge, but it is complex and time-consuming, requires high-level domain knowledge and is not easy to expand and maintain. Skeleton method forms an ontology skeleton by defining the concepts and relationships of goals, activities, resources, organizations and environment, which can be adjusted according to needs and is suitable for fields that need multi-perspective and multi-level knowledge, but it lacks formal semantics and reasoning ability. Based on this method, Ran et al. 34 constructed the ontology of idioms and allusions. IDEF5 method uses chart language and detailed description language to construct ontology, formalizes and visualizes objective knowledge, and is suitable for fields that need multi-source data and multi-participation, but it lacks a unified ontology representation language. Based on this method, Li et al. 35 constructed the business process activity ontology of military equipment maintenance support, and Song et al. 36 established the air defense and anti-missile operation process ontology. Methontology is a method close to software engineering. It systematically develops ontologies through the processes of specification, knowledge acquisition, conceptualization, integration, implementation, evaluation and document arrangement, which is suitable for fields that need multi-technology and multi-ontology integration, but it is too complicated and tedious, and requires a lot of resources and time 37 . Based on this method, Yang et al. 38 completed the ontology of emergency plan, Duan et al. 39 established the ontology of high-resolution images of rural residents, and Chen et al. 40 constructed the corpus ontology of Jiangui. Seven-step method is the most mature and widely used method at present 28 . It is systematic and applicable to construct ontology by determining its purpose, scope, terms, structure, attributes, limitations and examples 29 , but it does not provide quantitative indicators and methods about the quality and effect of ontology. Based on this method, Zhu et al. 41 constructed the disease ontology of asthma, Li et al. 42 constructed the ontology of military events, the ontology of weapons and equipment and the ontology model of battlefield environment, and Zhang et al. 43 constructed the ontology of stroke nursing field, and verified the construction results by expert consultation.

Domain ontology construction technology includes thesaurus conversion, existing ontology reuse and semi-automatic and automatic construction technology based on ontology engineering method 30 . The construction technology based on thesaurus transformation takes the existing thesaurus as the knowledge source, and transforms the concepts, terms and relationships in the thesaurus into the entities and relationships of domain ontology through certain rules and methods, which saves the time and cost of ontology construction and improves the quality and reusability of ontology. However, it is necessary to solve the structural and semantic differences between thesaurus and ontology and adjust and optimize them according to the characteristics of different fields and application scenarios. Wu et al. 44 constructed the ontology of the natural gas market according to the thesaurus of the natural gas market and the mapping of subject words to ontology, and Li et al. 45 constructed the ontology of the medical field according to the Chinese medical thesaurus. The construction technology based on existing ontology reuse uses existing ontologies or knowledge resources to generate new domain ontologies through modification, expansion, merger and mapping, which saves time and cost and improves the consistency and interoperability of ontologies, but it also needs to solve semantic differences and conflicts between ontologies. Chen et al. 46 reuse the top-level framework of scientific evidence source information ontology (SEPIO) and traditional Chinese medicine language system (TCMLS) to construct the ontology of clinical trials of traditional Chinese medicine, and Xiao et al. 47 construct the domain ontology of COVID-19 by extracting the existing ontology and the knowledge related to COVID-19 in the diagnosis and treatment guide. Semi-automatic and automatic construction technology based on ontology engineering method semi-automatically or automatically extracts the elements and structures of ontology from data sources by using natural language processing, machine learning and other technologies to realize large-scale, fast and low-cost domain ontology construction 48 , but there are technical difficulties, the quality and accuracy of knowledge extraction can not be well guaranteed, and the quality and consistency of different knowledge sources need to be considered. Suet al. 48 used regular templates and clustering algorithm to construct the ontology of port machinery, Zheng et al. 49 realized the automatic construction of mobile phone ontology through LDA and other models, Dong et al. 50 realized the automatic construction of ontology for human–machine ternary data fusion in manufacturing field, Linli et al. 51 proposed an ontology learning algorithm based on hypergraph, and Zhai et al. 52 learned from it through part-of-speech tagging, dependency syntax analysis and pattern matching.

At present, domain ontology construction methods are not easy to expand, lack of effective quality evaluation support, lack of effective integration and application of construction technology, construction divorced from reality can not guide subsequent practice, subjective ontology verification and so on. Aiming at the problems existing in the research of content characteristics and domain ontology construction of online rumors, this paper proposes an improved TFI network rumor domain ontology construction method based on seven-step method, which combines top-down existing ontology reuse technology with bottom-up semi-automatic construction technology, and establishes rumor domain ontology based on top-level ontology reuse, core document content feature extraction and new concept discovery in the real corpus from the terminology layer, framework layer and instance layer. Using Protégé as a visualization tool, the implicit knowledge mining of ontology is carried out by constructing SWRL rules to verify the semantic parsing ability and consistency of domain ontology.

Research method

This paper proposes a TFI online rumor domain ontology construction method based on the improvement of the seven-step method, which includes the term layer, the frame layer and the instance layer construction.

Term layer construction

Determine the domain and scope: the purpose of constructing the rumor domain ontology is to support the credible detection and governance of online rumors, and the domain and scope of the ontology are determined by answering questions.

Three-dimensional term set construction: investigate the top-level ontology and related core literature, complete the mapping of reusable top-level ontology and rumor content feature concept extraction semi-automatically from top to bottom; establish authoritative real rumor datasets, and complete the domain new concept discovery automatically from bottom to top; based on this, determine the term set of the domain ontology.

Frame layer construction

Define core classes and hierarchical relationships: combine the concepts of the three-dimensional rumor term set, based on the data distribution of the rumor dataset, define the parent class, summarize the subclasses, design hierarchical relationships and explain the content of each class.

Define core properties and facets of properties: in order to achieve deep semantic parsing of rumor text contents, define object properties, data properties and property facets for each category in the ontology.

Instance layer construction

Create instances: analyze the real rumor dataset, extract instance data, and add them to the corresponding concepts in the ontology.

Encode and visualize ontology: use OWL language to encode ontology, and use Protégé to visualize ontology, so that ontology can be understood and operated by computer.

Ontology verification: use SWRL rules and pellet reasoner to mine implicit knowledge of ontology, and verify its semantic parsing ability and consistency.

Ethical statements

This article does not contain any studies with human participants performed by any of the authors.

Determine the professional domain and scope of the ontology description

This paper determines the domain and scope of the online rumor domain ontology by answering the following four questions:

(1) What is the domain covered by the ontology?

The “Rumor Domain Ontology” constructed in this paper only considers content features, not user features and propagation features; the data covers six rumor types of politics and military, disease prevention and treatment, social life, science and technology, nutrition and health, and others involved in China’s mainstream internet rumor-refuting websites.

(2) What is the purpose of the ontology?

To perform fine-grained hierarchical modeling of the relationships among the features of multi-domain online rumor contents, realize semantic parsing and credibility reasoning verification of rumor texts, and guide fine-grained rumor detection and governance. It can also be used as a guiding framework and constraint condition for online rumor knowledge graph construction.

(3) What kind of questions should the information in the ontology provide answers for?

To provide answers for questions such as the fine-grained rumor types of rumor instances, the valid features of rumor types, etc.

(4) Who will use the ontology in the future?

Users of online rumor detection and governance, users of online rumor knowledge graphs construction.

Three-dimensional term set construction

Domain concepts reused by top-level ontology.

As a mature and authoritative common ontology, top-level ontology can be shared and reused in a large range, providing reference and support for the construction of domain ontology. The domain ontology of online rumors established in this paper focuses on the content characteristics, mainly including the content theme, events and emotions of rumor texts. By reusing the terminology concepts in the existing top-level ontology, the terminology in the terminology set can be unified and standardized. At the same time, the top-level concept and its subclass structure can guide the framework construction of domain ontology and reduce the difficulty and cost of ontology construction. Reusable top-level ontologies include: SUMO, senticnet and ERE after screening.

SUMO ontology: a public upper-level knowledge ontology containing some general concepts and relations for describing knowledge in different domains. The partial reusable SUMO top-level concepts and subclasses selected in this paper are shown in Table 1 , which provides support for the sub-concept design of text topics in rumor domain ontology.

Senticnet: a knowledge base for concept-based sentiment analysis, which contains semantic, emotional, and polarity information related to natural language concepts. The partial reusable SenticNet top-level concepts and subclasses selected in this paper are shown in Table 2 , which provides support for the sub-concept design of text topics in rumor domain ontology.

Entities, relations, and events (ERE): a knowledge base of events and entity relations. The partial reusable ERE top-level concepts and subclasses selected in this paper are shown in Table 3 , which provides support for the sub-concept design of text elements in the rumor domain ontology.

Extracting domain concepts based on core literature content features

Domain core literature is an important source for extracting feature concepts. This paper uses ‘rumor detection’ as the search term to retrieve 274 WOS papers and 257 CNKI papers from the WOS and CNKI core literature databases. The content features of rumor texts involved in the literature samples are extracted, the repetition content features are eliminated, the core content features are screened, and the canonical naming of synonymous concepts from different literatures yields the domain concepts as shown in Table 4 . Among them, text theme, text element, text style, text feature and text rhetoric are classified as text features; emotional category, emotional appeal and rumor motive are classified as emotional characteristics; source credibility, evidence credibility and testimony method are classified as information credibility characteristics; social context is implicit.

Extracting domain concepts based on new concept discovery

This paper builds a general rumor dataset based on China’s mainstream rumor-refuting websites as data sources, and proposes a domain new concept discovery algorithm to discover domain new words in the dataset, add them to the word segmentation dictionary to improve the accuracy of word segmentation, and cluster them according to rumor type, resulting in a concept subclass dictionary based on the real rumor dataset, which provided realistic basis and data support for the conceptual design of each subclass in domain ontology.

Building a general rumor dataset

The rumor dataset constructed in this paper contains 12,472 texts, with 6236 rumors and 6236 non-rumors; the data sources are China’s mainstream internet rumor-refuting websites: 1032 from the internet rumor exposure platform of China internet joint rumor-refuting platform, 270 from today’s rumor-refuting of China internet joint rumor-refuting platform, 1852 from Tencent news Jiaozhen platform, 1744 from Baidu rumor-refuting platform, 7036 from science rumor-refuting platform, and 538 from Weibo community management center. This paper invited eight researchers to annotate the labels (rumor, non-rumor), categories (politics and military, disease prevention and treatment, social life, science and technology, nutrition and health, others) of the rumor dataset. Because data annotation is artificial and subjective, in order to ensure the effectiveness and consistency of annotation, before inviting researchers to annotate, this paper formulates annotation standards, including the screening method, trigger words and sentence break identification of rumor information and corresponding rumor information, and clearly explains and exemplifies the screening method and trigger words of rumor categories, so as to reduce the understanding differences among researchers; in view of this standard, researchers are trained in labeling to familiarize them with labeling specifications, so as to improve their labeling ability and efficiency. The method of multi-person cross-labeling is adopted when labeling, and each piece of data is independently labeled by at least two researchers. In case of conflicting labeling results, the labeling results are jointly decided by the data annotators to increase the reliability and accuracy of labeling. After labeling, multi-person cross-validation method is used to evaluate the labeling results. Each piece of data is independently verified by at least two researchers who did not participate in labeling, and conflicting labeling results are jointly decided by at least five researchers to ensure the consistency of evaluation results. Examples of the results are shown in Table 5 .

N-gram word granularity rumor text new word discovery algorithm

Existing neologism discovery algorithms are mostly based on the granularity of Chinese characters, and the time complexity of long word discovery is high and the accuracy rate is low. The algorithm’s usefulness is low, and the newly discovered words are mostly already found in general domain dictionaries. To solve these problems, this paper proposes an online rumor new word discovery algorithm based on N-gram word granularity, as shown in Fig.  1 .

Flowchart of domain new word discovery algorithm.

First, obtain the corpus to be processed \({\varvec{c}}=\{{{\varvec{s}}}_{1},{{\varvec{s}}}_{2},...,{{\varvec{s}}}_{{{\varvec{n}}}_{{\varvec{c}}}}\}\) , and perform the first preprocessing on the corpus to be processed, which includes: sentence segmentation, Chinese word segmentation and punctuation removal for the corpus to be processed. Obtain the first corpus \({{\varvec{c}}}^{{\varvec{p}}}=\{{{\varvec{s}}}_{1}^{{\varvec{p}}},{{\varvec{s}}}_{2}^{{\varvec{p}}},...,{{\varvec{s}}}_{{{\varvec{n}}}_{{\varvec{c}}}}^{{\varvec{p}}}\}\) ; where \({s}_{i}\) represents the \(i\) -th sentence in the corpus to be processed, \({n}_{c}\) represents the number of sentences in the corpus to be processed, and \({s}_{i}^{p}\) is the i-th sentence in the first corpus; perform N-gram operation on each sentence in the first corpus separately, and obtain multiple candidate words \(n=2\sim 5\) ; count the word frequency of each candidate word in the first corpus, and remove the candidate words with word frequency less than the first threshold, and obtain the first class of candidate word set;calculate the cohesion of each candidate word in the first class of candidate word set according to the following formula:

In the formula, \(P(\cdot )\) represents word frequency.Then filter according to the second threshold corresponding to N-gram operation, and obtain the second class of candidate word set; after loading the new words in the second class of candidate word set into LTP dictionary, perform the second preprocessing on the corpus to be processed \({\varvec{c}}=\{{{\varvec{s}}}_{1},{{\varvec{s}}}_{2},...,{{\varvec{s}}}_{{{\varvec{n}}}_{{\varvec{c}}}}\}\) ; and obtain the second corpus \({{\varvec{c}}}^{{\varvec{p}}\boldsymbol{^{\prime}}}=\{{{\varvec{s}}}_{1}^{{\varvec{p}}\boldsymbol{^{\prime}}},{{\varvec{s}}}_{2}^{{\varvec{p}}\boldsymbol{^{\prime}}},...,{{\varvec{s}}}_{{{\varvec{n}}}_{{\varvec{c}}}}^{{\varvec{p}}\boldsymbol{^{\prime}}}\}\) ; where the second preprocessing includes: sentence segmentation, Chinese word segmentation and stop word removal for the corpus to be processed; after obtaining the vector representation of each word in the second corpus, determine the vector representation of each new word in the second class of candidate word set; according to the vector representation of each new word, use K-means algorithm for clustering; according to the clustering results and preset classification rules, classify each new word to the corresponding domain. The examples of new words discovered are shown in Table 6 :

RoBERTa-Kmeans rumor text concepts extraction algorithm

After adding the new words obtained by the new word discovery to the LTP dictionary, the accuracy of LTP word segmentation is improved. The five types of rumor texts established in this paper are segmented by using the new LTP dictionary, and the word vectors are obtained by inputting them into the RoBERTa word embedding layer after removing the stop words. The word vectors are clustered by k-means according to rumor type to obtain the concept subclass dictionary. The main process is as follows:

(1) Word embedding layer

The RoBERTa model uses Transformer-Encode for computation, and each module contains multi-head attention mechanism, residual connection and layer normalization, feed-forward neural network. The word vectors are obtained by representing the rumor texts after accurate word segmentation through one-hot encoding, and the position encoding represents the relative or absolute position of the word in the sequence. The word embedding vectors generated by superimposing the two are used as input X. The multi-head attention mechanism uses multiple independent Attention modules to perform parallel operations on the input information, as shown in formula ( 2 ):

where \(\left\{{\varvec{Q}},{\varvec{K}},{\varvec{V}}\right\}\) is the input matrix, \({{\varvec{d}}}_{{\varvec{k}}}\) is the dimension of the input matrix. After calculation, the hidden vectors obtained after computation are residual concatenated with layer normalization, and then calculated by two fully connected layers of feed-forward neural network for input, as shown in formula ( 3 ):

where \(\left\{{{\varvec{W}}}_{{\varvec{e}}},{{\varvec{W}}}_{0}\boldsymbol{^{\prime}}\right\}\) are the weight matrices of two connected layers, \(\left\{{{\varvec{b}}}_{{\varvec{e}}},{{\varvec{b}}}_{0}\boldsymbol{^{\prime}}\right\}\) are the bias terms of two connected layers.

After calculation, a bidirectional association between word embedding vectors is established, which enables the model to learn the semantic features contained in each word embedding vector in different contexts. Through fine-tuning, the learned knowledge is transferred to the downstream clustering task.

(2) K-means clustering

Randomly select k initial points to obtain k classes, and iterate until the loss function of the clustering result is minimized. The loss function can be defined as the sum of squared errors of each sample point from its cluster center point, as shown in formula ( 4 ).

where \({x}_{i}\) represents the \(i\) sample, \({a}_{i}\) is the cluster that \({x}_{i}\) belongs to, \({u}_{{a}_{i}}\) represents the corresponding center point, \(N\) is the total number of samples.

After RoBERTa-kmeans calculation, the concept subclasses obtained are manually screened, merged repetition items, deleted invalid items, and finally obtained 79 rumor concept subclasses, including 14 politics and military subclasses, 23 disease prevention and treatment subclasses, 15 social life subclasses, 13 science and technology subclasses, and 14 nutrition and health subclasses. Some statistics are shown in Table 7 .

Each concept subclass is obtained by clustering several topic words. For example, the topic words that constitute the subclasses of body part, epidemic prevention and control, chemical drugs, etc. under the disease prevention and treatment topic are shown in Table 8 .

(3) Determining the terminology set

This paper constructs a three-dimensional rumor domain ontology terminology set based on the above three methods, and unifies the naming of the terms. Some of the terms are shown in Table 9 .

Framework layer construction

Define core classes and hierarchy, define parent classes.

This paper aims at fine-grained hierarchical modeling of the relationship between the content characteristics of multi-domain network rumors. Therefore, the top-level parent class needs to include the rumor category and the main content characteristics of a sub-category rumor design. The main content characteristics are the clustering results of domain concepts extracted based on the content characteristics of core documents, that is, rumor text feature, rumor emotional characteristic, rumor credibility and social context. The specific contents of the five top parent classes are as follows:

Rumor type: the specific classification of rumors under different subject categories; Rumor text feature, the common features of rumor texts in terms of theme, style, rhetoric, etc. Rumor emotional characteristic: the emotional elements of rumor texts, the Rumor motive of the publisher, and the emotional changes they hope to trigger in the receiver. Rumor credibility: the authority of the information source, the credibility of the evidence material provided by the publisher, and the effectiveness of the testimony method. Social context: the relevant issues and events in the society when the rumor is published.

Induce subclasses and design hierarchical relationships

In this paper, under the top-level parent class, according to the top-level concepts of top-level ontologies such as SUMO, senticnet and ERE and their subclass structures, and the rumor text features of each category extracted from the real rumor text dataset, we summarize its 88 subclasses and design the hierarchical relationships, as shown in Fig.  2 , which include:

(1) Rumor text feature

Diagram of the core classes and hierarchy of the rumor domain ontology.

① Text theme 6 , 8 , 13 , 18 , 53 : the theme or topic that the rumor text content involves. Based on the self-built rumor dataset, it is divided into politics and military 54 , involving information such as political figures, political policies, political relations, political activities, military actions, military events, strategic objectives, politics and military reviews, etc.; nutrition and health 55 , involving information such as the relationship between human health and nutrition, the nutritional components and value of food, the plan and advice for healthy eating, health problems and habits, etc.; disease prevention and treatment 10 , involving information such as the definition of disease, vaccine, treatment, prevention, data, etc.; social life 56 , involving information such as social issues, social environment, social values, cultural activities, social media, education system, etc.; science and technology 57 , involving information such as scientific research, scientific discovery, technological innovation, technological application, technological enterprise, etc.; other categories.

② Text element 15 : the structured information of the rumor text contents. It is divided into character, political character, public character, etc.; geographical position, city, region, area, etc.; event, historical event, current event, crisis event, policy event, etc.; action, protection, prevention and control, exercise, fighting, crime, eating, breeding, health preservation, rest, exercise, education, sports, social, cultural, ideological, business, economic, transportation, etc.; material, food, products (food, medicine, health products, cosmetics, etc.) and the materials they contain and their relationship with human health. effect, nutrition, health, harm, natural disaster, man-made disaster, guarantee, prevention, treatment, etc.; institution, government, enterprise, school, hospital, army, police, social group, etc.; nature, weather, astronomy, environment, agriculture, disease, etc.

③ Text style 7 , 10 : the discourse style of the rumor text contents, preferring exaggerated and emotional expression. It is divided into gossip style, creating conflict or entertainment effect; curious style, satisfying people’s curiosity and stimulation; critical style, using receivers’ stereotypes or preconceptions; lyrical style, creating resonance and influencing emotion; didactic style influencing receivers’ thought and behavior from an authoritative perspective; plain style concise objective arousing resonance etc.

④ Text feature 7 , 58 : special language means in the rumor text contents that can increase the transmission and influence of the rumor. It is divided into extensive punctuation reminding or attracting receivers’ attention; many mood words enhancing emotional color and persuasiveness; many emoji conveying attitude; induce forwarding using @ symbol etc. to induce receivers to forward etc.

⑤ Text rhetoric 15 : common rhetorical devices in rumor contents. It is divided into metaphor hyperbole repetition personification etc.

(2) Rumor emotional characteristic

① Emotion category 17 , 59 , 60 : the emotional tendency and intensity expressed in the rumor texts. It is divided into positive emotion happy praise etc.; negative emotion fear 10 anger sadness anxiety 61 dissatisfaction depression etc.; neutral emotion no preference plain objective etc.

② Emotional appeal 16 , 62 , 63 : the online rumor disseminator hopes that the rumor they disseminate can trigger some emotional changes in the receiver. It is divided into “joy” happy pleasant satisfied emotions that prompt receivers to spread or believe some rumors that are conducive to social harmony; “love” love appreciation admiration emotions that prompt receivers to spread or believe some rumors that are conducive to some people or group interests; “anger” angry annoyed dissatisfied emotions that prompt receivers to spread or believe some rumors that are anti-social or intensify conflicts; “fear” fearful afraid nervous emotions that prompt receivers to spread or believe some rumors that have bad effects deliberately exaggerated; “repugnance” disgusted nauseous emotions that prompt receivers to spread or believe some rumors that are detrimental to social harmony; “surprise” surprised shocked amazed emotions that prompt receivers to spread or believe some rumors that deliberately attract traffic exaggerated fabricated etc.

③ Rumor motive 17 , 64 , 65 , 66 : the purpose and need of the rumor publisher to publish rumors and the receiver to forward rumors. Such as profit-driven seeking fame and fortune deceiving receivers; emotional catharsis relieving dissatisfaction emotions by venting; creating panic creating social unrest and riots disrupting social order; entertainment fooling receivers seeking stimulation; information verification digging out the truth of events etc.

(3) Rumor credibility

① source credibility 7 , 17 : the degree of trustworthiness that the information source has. Such as official institutions and authoritative experts and scholars in the field with high credibility; well-known encyclopedias and large-scale civil organizations with medium credibility; small-scale civil organizations and personal hearsay personal experience with low credibility etc.

② evidence credibility 61 : the credibility of the information proof material provided by the publisher. Data support such as scientific basis based on scientific theory or method; related feature with definite research or investigation result in data support; temporal background with clear time place character event and other elements which related to the information content; the common sense of life in line with the facts and scientific common sense that are widely recognized.

③ testimony method 10 , 11 , 17 : the method to support or refute a certain point of view. Such as multimedia material expressing or fabricating content details through pictures videos audio; authority endorsement policy documents research papers etc. of authorized institutions or persons; social identity identity of social relation groups.

(4) Social context

① social issue 67 : some bad phenomena or difficulties in society such as poverty pollution corruption crime government credibility decline 68 etc.

② public attention 63 : events or topics that arouse widespread attention or discussion in the society such as sports events technological innovation food safety religious beliefs Myanmar fraud nuclear wastewater discharge etc.

③ emergency(public sentiment) 69 : some major or urgent events that suddenly occur in society such as earthquake flood public safety malignant infectious disease outbreaks etc.

(5) Rumor type

① Political and military rumor:

Political image rumor: rumors related to images closely connected to politics and military, such as countries, political figures, institutions, symbols, etc. These include positive political image smear rumor, negative political image whitewash rumor, political image fabrication and distortion rumor, etc.

Political event rumor: rumors about military and political events, such as international relations, security cooperation, military strategy, judicial trial, etc. These include positive political event smear rumor, negative political event whitewash rumor, political event fabrication and distortion rumor, etc.

② Nutrition and health rumor:

Food product rumor: rumors related to food, products (food, medicine, health products, cosmetics, etc.), the materials they contain and their association with human health. These include positive effect of food product rumor, negative effect of food product rumor, food product knowledge rumor, etc.

Living habit rumor: rumors related to habitual actions in life and their association with human health. These include positive effect of living habit rumor, negative effect of living habit rumor, living habit knowledge rumor, etc.

③ Disease prevention and treatment rumor:

Disease management rumor: rumors related to disease management and control methods that maintain and promote individual and group health. These include positive prevention and treatment rumor, negative aggravating disease rumor, disease management knowledge rumor, etc.

Disease confirmed transmission rumor: rumors about the confirmation, transmission, and immunity of epidemic diseases at the social level in terms of causes, processes, results, etc. These include local confirmed cases rumor, celebrity confirmed cases rumor, transmission mechanism rumor, etc.

Disease notification and advice rumor: rumors that fabricate or distort the statements of authorized institutions or experts in the field, and provide false policies or suggestions related to diseases. These include institutional notification rumor, expert advice rumor, etc.

④ Social life rumor:

Public figure public opinion rumor: rumors related to public figures’ opinions, actions, private lives, etc. These include positive public figure smear rumor, negative public figure whitewash rumor, public figure life exposure rumor, etc.

Social life event rumor: rumors related to events, actions, and impacts on people's social life. These include positive event sharing rumor, negative event exposure rumor, neutral event knowledge rumor, etc.

Disaster occurrence rumor: rumors related to natural disasters or man-made disasters and their subsequent developments. These include natural disaster occurrence rumor, man-made disaster occurrence rumor, etc.

⑤ Science and technology rumor:

Scientific knowledge rumor: rumors related to natural science or social science theories and knowledge. These include scientific theory rumor, scientific concept rumor, etc.

Science and technology application rumor: rumors related to the research and development and practical application of science and technology and related products. These include scientific and technological product rumor, scientific and technological information rumor, etc.

⑥ Other rumor: rumors that do not contain elements from the above categories.

Definition of core properties and facets of properties

Properties in the ontology are used to describe the relationships between entities or the characteristics of entities. Object properties are relationships that connect two entities, describing the interactions between entities; data properties represent the characteristics of entities, usually in the form of some data type. Based on the self-built rumor dataset, this paper designs object properties, data properties and facets of properties for the parent classes and subclasses of the rumor domain ontology.

Object properties

A partial set of object properties is shown in Table 10 .

Data attributes

The partial data attribute set is shown in Table 11 .

Creating instances

Based on the defined core classes and properties, this paper creates instances according to the real rumor dataset. An example is shown in Table 12 .

This paper selects the online rumor that “Lin Chi-ling was abused by her husband Kuroki Meisa, the tears of betrayal, the shadow of gambling, all shrouded her head. Even if she tried to divorce, she could not get a solution…..” as an example, and draws a structure diagram of the rumor domain ontology instance, as shown in Fig.  3 . This instance shows the seven major text features of the rumor text: text theme, text element, text style, emotion category, emotional appeal, rumor motivation, and rumor credibility, as well as the related subclass instances, laying a foundation for building a multi-source rumor domain knowledge graph.

Schematic example of the rumor domain ontology.

Encoding ontology and visualization

Encoding ontology.

This paper uses OWL language to encode the rumor domain ontology, to accurately describe the entities, concepts and their relationships, and to facilitate knowledge reasoning and semantic understanding. Classes in the rumor domain ontology are represented by the class “Class” in OWL and the hierarchical relationship is represented by subclassof. For example, in the creation of the rumor emotional characteristic class and its subclasses, the OWL code is shown in Fig.  4 :

Partial OWL codes of the rumor domain ontology.

The ontology is formalized and stored as a code file using the above OWL language, providing support for reasoning.

Ontology visualization

This paper uses protégé5.5 to visualize the rumor domain ontology, showing the hierarchical structure and relationship of the ontology parent class and its subclasses. Due to space limitations, this paper only shows the ontology parent class “RumorEmotionalFeatures” and its subclasses, as shown in Fig.  5 .

Ontology parent class “RumorEmotionalFeatures” and its subclasses.

Ontology reasoning and validation

Swrl reasoning rule construction.

SWRL reasoning rule is an ontology-based rule language that can be used to define Horn-like rules to enhance the reasoning and expressive ability of the ontology. This paper uses SWRL reasoning rules to deal with the conflict relationships between classes and between classes and instances in the rumor domain ontology, and uses pellet reasoner to deeply mine the implicit semantic relationships between classes and instances, to verify the semantic parsing ability and consistency of the rumor domain ontology.

This paper summarizes the object property features of various types of online rumors based on the self-built rumor dataset, maps the real rumor texts with the rumor domain ontology, constructs typical SWRL reasoning rules for judging 32 typical rumor types, as shown in Table 13 , and imports them into the protégé rule library, as shown in Fig.  6 . In which x, n, e, z, i, t, v, l, etc. are instances of rumor types, text theme, emotion category, effect, institution, event, action, geographical position, etc. in the ontology. HasTheme, HasEmotion, HasElement, HasSource, HasMood and HasSupport are object property relationships. Polarity value is a data property relationship.

Partial SWRL rules for the rumor domain ontology.

Implicit knowledge mining and verification based on pellet reasoner

This paper extracts corresponding instances from the rumor dataset, imports the rumor domain ontology and SWRL rule description into the pellet reasoner in the protégé software, performs implicit knowledge mining of the rumor domain ontology, judges the rumor type of the instance, and verifies the semantic parsing ability and consistency of the ontology.

Positive prevention and treatment of disease rumors are mainly based on the theme of disease prevention and treatment, usually containing products to be sold (including drugs, vaccines, equipment, etc.) and effect of disease names, claiming to have positive effects (such as prevention, cure, relief, etc.) on certain diseases or symptoms, causing positive emotions such as surprise and happiness among patients and their families, thereby achieving the purpose of selling products. The text features and emotional features of this kind of rumors are relatively clear, so this paper takes the rumor text “Hong Kong MDX Medical Group released the ‘DCV Cancer Vaccine’, which can prevent more than 12 kinds of cancers, including prostate cancer, breast cancer and lung cancer.” as an example to verify the semantic parsing ability of the rumor domain ontology. The analysis result of this instance is shown in Fig.  7 . The text theme is cancer prevention in disease prevention and treatment, the text style is plain narrative style, and the text element includes product-DCV cancer vaccine, positive effect-prevention, disease name-prostate cancer, disease name-breast cancer, disease name-lung cancer; the emotion category of this instance is a positive emotion, emotional appeal is joy, love, surprise; The motive for releasing rumors is profit-driven in selling products, the information source is Hong Kong MDX medical group, and pictures and celebrity endorsements are used as testimony method. This paper uses a pellet reasoner to reason on the parsed instance based on SWRL rules, and mines out the specific rumor type of this instance as positive prevention and treatment of disease rumor. This paper also conducted similar instance analysis and reasoning verification for other types of rumor texts, and the results show that the ontology has high consistency and reliability.

Implicit relationship between rumor instance parsing results and pellet reasoner mining.

Comparison and evaluation of ontology performance

In this paper, the constructed ontology is compared with the representative rumor index system in the field. By inviting four experts to make a comprehensive evaluation based on the self-built index system 70 , 71 , 72 , their performance in the indicators of reliability, coverage and operability is evaluated. According to the ranking order given by experts, they are given 1–4 points, and the first place in each indicator item gets four points. The average value given by three experts is taken as the single indicator score of each subject, and the total score of each indicator item is taken as the final score of the subject.

As can be seen from Table 14 , the rumor domain ontology constructed in this paper constructs a term set through three ways: reusing the existing ontology, extracting the content features of core documents and discovering new concepts based on real rumor data sets, and the ontology structure has been verified by SWRL rule reasoning of pellet inference machine, which has high reliability; ontology covers six kinds of Chinese online rumors, including the grammatical, semantic, pragmatic and social characteristics of rumor text characteristics, emotional characteristics, rumor credibility and social background, which has a high coverage; ontology is coded by OWL language specification and displayed visually on protege, which is convenient for further expansion and reuse of scholars and has high operability.

The construction method of TFI domain ontology proposed in this paper includes terminology layer, framework layer and instance layer. Compared with the traditional methods, this paper adopts three-dimensional data set construction method in terminology layer construction, investigates top-level ontology and related core documents, and completes the mapping of reusable top-level ontology from top to bottom and the concept extraction of rumor content features in existing literature research. Based on the mainstream internet rumor websites in China, the authoritative real rumor data set is established, and the new word discovery algorithm of N-gram combined with RoBERTa-Kmeans clustering algorithm is used to automatically discover new concepts in the field from bottom to top; determine the terminology set of domain ontology more comprehensively and efficiently. This paper extracts the clustering results of domain concepts based on the content characteristics of core documents in the selection of parent rumors content characteristics in the framework layer construction, that is, rumors text characteristics, rumors emotional characteristics, rumors credibility characteristics and social background characteristics; based on the emotional characteristics and the entity categories of real rumor data sets, the characteristics of rumor categories are defined. Sub-category rumor content features combine the concept of three-dimensional rumor term set and the concept distribution based on real rumor data set, define the sub-category concept and hierarchical relationship close to the real needs, and realize the fine-grained hierarchical modeling of the relationship between multi-domain network rumor content features. In this paper, OWL language is used to encode the rumor domain ontology in the instance layer construction, and SWRL rule language and Pellet inference machine are used to deal with the conflict and mine tacit knowledge, judge the fine-grained categories of rumor texts, and realize the effective quality evaluation of rumor ontology. This makes the rumor domain ontology constructed in this paper have high consistency and reliability, and can effectively analyze and reason different types of rumor texts, which enriches the knowledge system in this field and provides a solid foundation for subsequent credible rumor detection and governance.

However, the study of the text has the following limitations and deficiencies:

(1) The rumor domain ontology constructed in this paper only considers the content characteristics, but does not consider the user characteristics and communication characteristics. User characteristics and communication characteristics are important factors affecting the emergence and spread of online rumors, and the motivation and influence of rumors can be analyzed. In this paper, these factors are not included in the rumor feature system, which may limit the expressive ability and reasoning ability of the rumor ontology and fail to fully reflect the complexity and multidimensional nature of online rumors.

(2) In this paper, the mainstream Internet rumor-dispelling websites in China are taken as the data source of ontology instantiation. The data covers five rumor categories: political and military, disease prevention, social life, science and technology, and nutrition and health, and the data range is limited. And these data sources are mainly official or authoritative rumor websites, and their data volume and update frequency may not be enough to reflect the diversity and variability of online rumors, and can not fully guarantee the timeliness and comprehensiveness of rumor data.

(3) The SWRL reasoning rules used in this paper are based on manual writing, which may not cover all reasoning scenarios, and the degree of automation needs to be improved. The pellet inference engine used in this paper is an ontology inference engine based on OWL-DL, which may have some computational complexity problems and lack of advanced reasoning ability.

The following aspects can be considered for optimization and improvement in the future:

(1) This paper will introduce user characteristics into the rumor ontology, and analyze the factors that cause and accept rumors, such as social attributes, psychological state, knowledge level, beliefs and attitudes, behavioral intentions and so on. This paper will introduce the characteristics of communication, and analyze the propagation dynamic factors of various types of rumors, such as propagation path, propagation speed, propagation range, propagation period, propagation effect, etc. This paper hopes to introduce these factors into the rumor feature system, increase the breadth and depth of the rumor domain ontology, and provide more credible clues and basis for the detection, intervention and prevention of rumors.

(2) This paper will expand the data sources, collect the original rumor data directly from social media, news media, authoritative rumor dispelling institutions and other channels, and build a rumor data set with comprehensive types, diverse expressions and rich characteristics; regularly grab the latest rumor data from these data sources and update and improve the rumor data set in time; strengthen the expressive ability of rumor ontology instance layer, and provide full data support and verification for the effective application of ontology.

(3) The text will introduce GPT, LLaMA, ChantGLM and other language models, and explore the automatic generation algorithm and technology of ontology inference rules based on rumor ontology and dynamic Prompt, so as to realize more effective and intelligent rumor ontology evaluation and complex reasoning.

This paper proposed a method of constructing TFI network rumor domain ontology. Based on the concept distribution of three-dimensional term set and real rumor data set, the main features of network rumors are defined, including text features, emotional features, credibility features, social background features and category features, and the relationships among these multi-domain features are modeled in a fine-grained hierarchy, including five parent classes and 88 subcategories. At the instance level, 32 types of typical rumor category judgment and reasoning rules are constructed, and the ontology is processed by using SWRL rule language and pellet inference machine for conflict processing and tacit knowledge mining, so that the semantic analysis and reasoning of rumor text content are realized, which proves its effectiveness in dealing with complex, fuzzy and uncertain information in online rumors and provides a new perspective and tool for the interpretable analysis and processing of online rumors.

Data availability

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Jiang, S. The production scene and content characteristics of scientific rumors. Youth J. https://doi.org/10.15997/j.cnki.qnjz.2020.33.011 (2020).

Article   Google Scholar  

Jin, X. & Zhao, Y. Analysis of internet rumors from the perspective of co-governance—Practice of rumor governance on wechat platform. News and Writing. 6 , 41–44 (2017).

Bai, S. Research on the causes and countermeasures of internet rumors. Press https://doi.org/10.15897/j.cnki.cn51-1046/g2.2010.04.035 (2010).

Garg, S. & Sharma, D. K. Linguistic features based framework for automatic fake news detection. Comput. Ind. Eng. 172 , 108432 (2022).

Zhao, J., Fu, C. & Kang, X. Content characteristics predict the putative authenticity of COVID-19 rumors. Front. Public Health 10 , 920103 (2022).

Article   PubMed   PubMed Central   Google Scholar  

Zhang, Z., Shu, K. & He, L. The theme and characteristics of wechat rumors. News and Writing. 1 , 60–64 (2016).

Li, B. & Yu, G. Research on the discourse space and communication field of internet rumors in the post-truth era—Based on the analysis of 4160 rumors in wechat circle of friends. Journalism Research. 2 , 103–112 (2018).

Yu, G. Text structure and expression characteristics of internet rumors—Analysis of 6000+ rumors based on tencent big data screening and identification. News and Writing. 2 , 53–59 (2018).

Mourão, R. R. & Robertson, C. T. Fake news as discursive integration: An analysis of sites that publish false, misleading, hyperpartisan and sensational information. J. Stud. 20 , 2077–2095 (2019).

Google Scholar  

Zhou, G. Analysis on the content characteristics and strategies of epidemic rumors—Based on Sina’s “novel coronavirus epidemic rumors list”. Sci. Popul. https://doi.org/10.19293/j.cnki.1673-8357.2021.05.002 (2021).

Huang, Y. An analysis of the internal logic and methods of rumor “confirmation”—An empirical study based on 60 rumors spread on wechat. J. Party Sch. Tianjin Munic. Comm. CPC 20 , 7 (2018).

Butt, S. et al . What goes on inside rumour and non-rumour tweets and their reactions: A psycholinguistic analyses. Comput. Hum. Behav. 135 , 107345 (2022).

Zhou, L., Tao, J. & Zhang, D. Does fake news in different languages tell the same story? An analysis of multi-level thematic and emotional characteristics of news about COVID-19. Inf. Syst. Front. 25 , 493–512. https://doi.org/10.1007/s10796-022-10329-7 (2023).

Article   PubMed   Google Scholar  

Tan, L. et al . Research status of deep learning methods for rumor detection. Multimed. Tools Appl. 82 , 2941–2982 (2023).

Damstra, A. et al. What does fake look like? A review of the literature on intentional deception in the news and on social media. J. Stud. 22 , 1947–1963. https://doi.org/10.1080/1461670X.2021.1979423 (2021).

Lai, S. & Tang, X. Research on the influence of information emotionality on the spread of online rumors. J. Inf. 35 , 116–121 (2016).

ADS   Google Scholar  

Yuan, H. & Xie, Y. Research on the rumor maker of internet rumors about public events—Based on the content analysis of 118 influential Internet rumors about public events. Journalist https://doi.org/10.16057/j.cnki.31-1171/g2.2015.05.008 (2015).

Ruan, Z. & Yin, L. Types and discourse focus of weibo rumors—Based on the content analysis of 307 weibo rumors. Contemporary Communication. 4 , 77–78+84 (2014).

Zhang, W. & Zhu, Q. Research on the Construction Method of Domain Ontology. Books and Information. 5 , 16–19+40 (2011).

Tham, K.D., Fox, M.S. & Gruninger, M. A cost ontology for enterprise modelling. In Proceedings of 3rd IEEE Workshop on Enabling Technologies: Infrastructure for Collaborative Enterprises. IEEE , 197–210. https://doi.org/10.1109/ENABL.1994.330502 (1994).

Uschold, M. & Gruninger, M. Ontologies: Principles, methods and applications. Knowl. Eng. Rev. 11 , 93–136 (1996).

Menzel, C. P., Mayer, R. J. & Painter, M. K. IDEF5 ontology description capture method: Concepts and formal foundations (Armstrong Laboratory, Air Force Materiel Command, Wright-Patterson Air Force, 1992).

Book   Google Scholar  

Song, Z., Zhu, F. & ZHANG, D. Research on air and missile defense domain ontology development based on IDEF5 and OWL. Journal of Projectiles, Rockets, Missiles and Guidance. 30 , 176–178 (2010).

Fernández-López, M., Gómez-Pérez, A. & Juristo, N. Methontology: From ontological art towards ontological engineering. AAAI-97 Spring Symposium Series . https://oa.upm.es/5484/ (1997).

Sawsaa, A. & Lu, J. Building information science ontology (OIS) with methontology and protégé. J. Internet Technol. Secur. Trans. 1 , 100–109 (2012).

Yue, L. & Liu, W. Comparative study on the construction methods of domain ontology at home and abroad. Inf. Stud. Theory Appl. 39 , 119–125. https://doi.org/10.16353/j.cnki.1000-7490.2016.08.024 (2016).

Noy, N.F. & McGuinness, D.L. Ontology development 101: A guide to creating your first ontology. Stanford knowledge systems laboratory technical report. KSL-01–05 (2001).

Luo, Y. et al . vim: Research on OWL-based vocabulary ontology construction method for units of measurement. Electronics 12 , 3783 (2023).

Al-Aswadi, F. N., Chan, H. Y. & Gan, K. H. Automatic ontology construction from text: A review from shallow to deep learning trend. Artif. Intell. Rev. 53 , 3901–3928 (2020).

Chen, X. & Mao, T. Ontology construction of documentary heritage—Taking China archives documentary heritage list as an example. Libr. Trib. 43 , 120–131 (2023).

CAS   Google Scholar  

Zhao, X. & Li, T. Research on the ontology construction of archives oriented to digital humanism—Taking Wanli tea ceremony archives as an example. Inf. Stud. Theory Appl. 45 , 154–161. https://doi.org/10.16353/j.cnki.1000-7490.2022.08.021 (2022).

Huang, X. et al . Construction of special knowledge base of government website pages based on domain ontology—Taking “COVID-19 vaccine science popularization” as an example. Libr. Inf. Serv. 66 , 35–46. https://doi.org/10.13266/j.issn.0252-3116.2022.17.004 (2022).

Jindal, R., Seeja, K. & Jain, S. Construction of domain ontology utilizing formal concept analysis and social media analytics. Int. J. Cogn. Comput. Eng. 1 , 62–69 (2020).

Ran, J. et al . Research on ontology construction of idioms and allusions based on OWL. Comput. Technol. Dev. 20 , 63–66 (2010).

Li, L. et al . Research on business process modeling of army equipment maintenance support based on IDEF5. Technol. Innov. Appl. 11 , 80–82 (2021).

Song, Z. et al . Ontology modeling of air defense and anti-missile operation process based on IDEF5/OWL. J. Missiles Guid. 30 , 176–178 (2010).

Li, A., Xu, Y. & Chi, Y. Summary of ontology construction and application. Inf. Stud. Theory Appl 46 , 189–195. https://doi.org/10.16353/j.cnki.1000-7490.2023.11.024 (2023).

Yang, J., Song, C. & Jin, L. Ontology construction of emergency plan based on methontology method. J. Saf. Environ. 18 , 1427–1431. https://doi.org/10.13637/j.issn.1009-6094.2018.04.033 (2018).

Duan, L. & Li, H. Ontology modeling method of high-resolution image rural residential area supported by OIA technology. Modern Agricultural Science and Technology. 2 , 338–340 (2016).

Chen, Y. & Jiang, H. Construction of fire inspection knowledge map based on GIS geospatial relationship. J. Subtrop. Resour. Environ. 18 , 109–118. https://doi.org/10.19687/j.cnki.1673-7105.2023.03.014 (2023).

Zhu, L. et al. Construction of TCM asthma domain ontology. Chin. J. Exp. Tradit. Med. Formulae 23 , 222–226. https://doi.org/10.13422/j.cnki.syfjx.2017150222 (2017).

Li, H. et al . Domain ontology construction and relational reasoning. J. Inf. Eng. Univ. 24 , 321–327 (2023).

Zhang, Y. et al. Construction of ontology of stroke nursing field based on corpus. Chin. Nurs. Res. 36 , 4186–4190 (2022).

Wu, M. et al. Ontology construction of natural gas market knowledge map. Pet. New Energy 34 , 71–76 (2022).

Li, X. et al . Research on ontology construction based on thesaurus and its semantic relationship. Inf. Sci. 36 , 83–87 (2018).

Article   ADS   CAS   Google Scholar  

Chen, Q. et al . Construction of knowledge ontology of clinical trial literature of traditional Chinese medicine. Chin. J. Exp. Tradit. Med. Formulae 29 , 190–197. https://doi.org/10.13422/j.cnki.syfjx.20231115 (2023).

Xiao, Y. et al. Construction and application of novel coronavirus domain ontology. Mil. Med. 46 , 263–268 (2022).

Su, N. et al . Automatic construction method of domain-limited ontology. Lifting the Transport Machinery. 8 , 49–57 (2023).

Zheng, S. et al . Ontology construction method for user-generated content. Inf. Sci. 37 , 43–47. https://doi.org/10.13833/j.issn.1007-7634.2019.11.007 (2019).

Dong, J., Wang, J. & Wang, Z. Ontology automatic construction method for human-machine-object ternary data fusion in manufacturing field. Control Decis. 37 , 1251–1257. https://doi.org/10.13195/j.kzyjc.2020.1298 (2022).

Zhu, L., Hua, G. & Gao, W. Mapping ontology vertices to a line using hypergraph framework. Int. J. Cogn. Comput. Eng. 1 , 1–8 (2020).

Zhai, Y. & Wang, F. Research on the construction method of Chinese domain ontology based on text mining. Inf. Sci. 33 , 3–10. https://doi.org/10.13833/j.cnki.is.2015.06.001 (2015).

Duan, Z. Generation mechanism of internet rumors and countermeasures. Guizhou Soc. Sci. https://doi.org/10.13713/j.cnki.cssci.2016.04.014 (2016).

Du, Z. & Zhi, S. The harm and governance of network political rumors. Academic Journal of Zhongzhou. 4 , 161–165 (2019).

Song, X. et al . Research on influencing factors of health rumor sharing willingness based on MOA theory. J. China Soc. Sci. Tech. Inf. 39 , 511–520 (2020).

Jiang, S. Research on the characteristics, causes and countermeasures of social rumors dissemination in china in recent years. Red Flag Manuscript . 16 , 4 (2011).

Huang, J., Wang, G. & Zhong, S. Research on the propagation law and function mode of sci-tech rumors. Journal of Information. 34 , 156–160 (2015).

Liu, Y. et al . A survey of rumor recognition in social media. Chin. J. Comput. 41 , 1536–1558 (2018).

Wei, D. et al. Public emotions and rumors spread during the covid-19 epidemic in China: Web-based correlation study. J. Med. Internet Res. 22 , e21933 (2020).

Runxi, Z. & Di, Z. A model and simulation of the emotional contagion of netizens in the process of rumor refutation. Sci. Rep. https://doi.org/10.1038/s41598-019-50770-4 (2019).

Tang, X. & Lai, S. Research on the forwarding of network health rumors in public health security incidents—Interaction between perceived risk and information credibility. J. Inf. 40 , 101–107 (2021).

Nicolas, P., Dominik, B. & Stefan, F. Emotions in online rumor diffusion. EPJ Data Sci. https://doi.org/10.1140/epjds/s13688-021-00307-5 (2021).

Deng, G. & Tang, G. Research on the spread of network rumors and its social impact. Seeker https://doi.org/10.16059/j.cnki.cn43-1008/c.2005.10.031 (2005).

Ji, Y. Research on the communication motivation of wechat rumors. Youth J. https://doi.org/10.15997/j.cnki.qnjz.2019.17.006 (2019).

Yuan, G. Analysis on the causes and motives of internet rumors in emergencies—Taking social media as an example. Media. 21 , 80–83 (2016).

Zhao, N., Li, Y. & Zhang, J. A review of the research on influencing factors and motivation mechanism of rumor spread. J. Psychol. Sci. 36 , 965–970. https://doi.org/10.16719/j.cnki.1671-6981.2013.04.015 (2013).

Article   CAS   Google Scholar  

Hu, H. On the formation mechanism of social rumors from the perspective of “rumors and salt storm”. J. Henan Univ. 52 , 63–68 (2012).

Yue, Y. et al. Trust in government buffers the negative effect of rumor exposure on people’s emotions. Curr. Psychol. 42 , 23917–23930 (2023).

Wang, C. & Hou, X. Analysis of rumor discourse in major emergencies. J. Commun. 19 , 34–38 (2012).

Xu, L. Research progress of ontology evaluation. J. China Soc. Scie. Tech. Inf. 35 , 772–784 (2016).

Lantow, B. & Sandkuhl, K. An analysis of applicability using quality metrics for ontologies on ontology design patterns. Intell. Syst. Acc. Financ. Manag. 22 , 81–99 (2015).

Pak, J. & Zhou, L. A framework for ontology evaluationIn. Exploring the Grand Challenges for Next Generation E-Business: 8th Workshop on E-Business, WEB 2009, Phoenix, AZ, USA, December 15, 2009, Revised Selected Papers 8. , 10–18. https://doi.org/10.1007/978-3-642-17449-0_2 (Springer Berlin Heidelberg, 2011).

Download references

Acknowledgements

This study was financially supported by Xi'an Major Scientific and Technological Achievements Transformation and Industrialization Project (20KYPT0003-10).

This work was supported by Xi’an Municipal Bureau of Science and Technology, 20KYPT0003-10.

Author information

Authors and affiliations.

School of Economics and Management, Xidian University, 266 Xifeng Road, Xi’an, 710071, China

Jianbo Zhao, Huailiang Liu, Weili Zhang, Tong Sun, Qiuyi Chen, Yan Zhuang, Xiaojin Zhang & Shanzhuang Zhang

School of Artificial Intelligence, Xidian University, 266 Xifeng Road, Xi’an, 710071, China

Yuehai Wang, Jiale Cheng & Ruiyu Ding

School of Telecommunications Engineering, Xidian University, 266 Xifeng Road, Xi’an, 710071, China

You can also search for this author in PubMed   Google Scholar

Contributions

H.L. formulated the overall research strategy and guided the work. J.Z kept the original data on which the paper was based and verified whether the charts and conclusions accurately reflected the collected data. J.Z. W.Z. and T.S. wrote the main manuscript text. W.Z. Y.W. and Q.C. finished collecting and sorting out the data. J.C. Y.Z. and X.Z. prepared Figs.  1 – 7 , S.Z. B.L. and R.D. prepared Tables 1 – 14 . All authors reviewed the manuscript.

Corresponding author

Correspondence to Jianbo Zhao .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Zhao, J., Liu, H., Zhang, W. et al. Research on domain ontology construction based on the content features of online rumors. Sci Rep 14 , 12134 (2024). https://doi.org/10.1038/s41598-024-62459-4

Download citation

Received : 07 December 2023

Accepted : 16 May 2024

Published : 27 May 2024

DOI : https://doi.org/10.1038/s41598-024-62459-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Rumor content features
• Domain ontology
• Top-level ontology reuse
• New concept discovery

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.

• Reference Manager
• Simple TEXT file

People also looked at

Technology and code article, a high-precision interpretable framework for marine dissolved oxygen concentration inversion.

• 1 College of Computer and Communication Engineering, China University of Petroleum, Qingdao, China
• 2 School of Artificial Intelligence, Nanjing University of Information Science andTechnology, Nanjing, China

Variations in Marine Dissolved Oxygen Concentrations (MDOC) play a critical role in the study of marine ecosystems and global climate evolution. Although artificial intelligence methods, represented by deep learning, can enhance the precision of MDOC inversion, the uninterpretability of the operational mechanism involved in the “black-box” often make the process difficult to interpret. To address this issue, this paper proposes a high-precision interpretable framework (CDRP) for intelligent MDOC inversion, including Causal Discovery, Drift Detection, RuleFit Model, and Post Hoc Analysis. The entire process of the proposed framework is fully interpretable: (i) The causal relationships between various elements are further clarified. (ii) During the phase of concept drift analysis, the potential factors contributing to changes in marine data are extracted. (iii) The operational rules of RuleFit ensure computational transparency. (iv) Post hoc analysis provides a quantitative interpretation from both global and local perspectives. Furthermore, we have derived quantitative conclusions about the impacts of various marine elements, and our analysis maintains consistency with conclusions in marine literature on MDOC. Meanwhile, CDRP also ensures the precision of MDOC inversion: (i) PCMCI causal discovery eliminates the interference of weakly associated elements. (ii) Concept drift detection takes more representative key frames. (iii) RuleFit achieves higher precision than other models. Experiments demonstrate that CDRP has reached the optimal level in single point buoy data inversion task. Overall, CDRP can enhance the interpretability of the intelligent MDOC inversion process while ensuring high precision.

1 Introduction

Marine Dissolved Oxygen Concentration (MDOC) serves as an essential indicator for evaluating seawater conditions and plays a significant role in the regulation of the global climate. The decrease in MDOC, also known as marine hypoxia, can significantly affect marine ecosystems, potentially leading to extensive marine biota mortality events ( Karadurmus and Sari, 2022 ; Brock et al., 2023 ; Wang et al., 2023b ). This phenomenon directly impacts 10% to 12% of the global population reliant on coastal ecosystems for sustenance ( Breitburg et al., 2018 ; Li et al., 2023b ). On the other hand, the production of nitrous oxide (N 2 O) shows obvious sensitivity to variations in MDOC. Particularly in conditions of reduced MDOC, there is a notable rise in N 2 O production ( Suntharalingam et al., 2000 ; Jin and Gruber, 2003 ; Hutchins and Capone, 2022 ). Despite the importance of MDOC research within the field of marine science, the accessibility of MDOC data remains relatively constrained in comparison to data on temperature and salinity. This limitation hinders comprehensive research efforts in this area ( Wang et al., 2020 ). Currently, widely used marine data include buoy measurements of MDOC and other marine elements from the World’s Oceans Real-time Network Plan (ARGO) 1 , as well as the World Ocean Database (WOD) 2 , which compiles datasets from various countries and organizations. However, initial deployments primarily focused on measuring temperature and salinity, meanwhile modern buoys face challenges related to calibration and drift ( Johnson et al., 2017 ). Consequently, utilizing data such as temperature and salinity to infer MDOC holds great significance, making MDOC inversion highly meaningful.

The development of MDOC inversion methodologies has primarily undergone three stages: numerical computation, machine learning, and deep learning ( Figure 1 ): Initially, numerical computation was employed for inversion calculations, but the associated computational costs were found to be excessively high; Nonetheless, the introduction of machine learning methodologies within the domain of artificial intelligence significantly reduced computational costs ( Figure 1A ); More recently, the deployment of deep learning models has further elevated computational precision, but the uninterpretable working mechanism of the “black-box” has led to the low interpretaiblity of these models ( Figure 1B ); After conducting research, we found that rule-based methods such as RuleFit, excelling in both precision and interpretability, have not been widely applied in oceanography, thus their adoption could effectively ensure high precision and inherent interpretability in marine intelligent inversion models ( Figure 1C ).

Figure 1 Precision and interpretability in MDOC inversion tasks. (A) Machine Learning; (B) Deep Learning; (C) Rule-based.

At present, there have been multiple approaches to address the MDOC inversion. Traditionally, using climate system models and low-order marine biogeochemical models for MDOC inversion has been a common practice ( Matear and Hirst, 2003 ), while mathematical modeling is also a prevalent method for MDOC inversion ( Naik and Manjappa, 2011 ). However, these traditional models have some limitations, such as slow computational speed, demanding equipment requirements, and high operational costs, making it difficult to implement streamlined inversion for MDOC.

Nowadays with the rapid expansion of marine datasets, machine learning has supassed traditional methods in robustness and has shown excellent performance in uncovering the complex nonlinear relationships between variables ( Jiang et al., 2017 ), because of its faster computational speeds and lower dependence on data assumptions. And multiple machine learning algorithms have been emplyed to investigate the association between dissolved oxygen concentration and other elements. Ji et al. (2017) utilized eleven hydrochemical variables from the Wen-Rui Tang River to assess the accuracy of dissolved oxygen concentration inversion using Support Vector Regression (SVR). Giglio et al. (2018) attempted to use Random Forest Regression (RFR) to reproduce the dissolved oxygen concentration fields from the Southern Ocean State Estimate (SOSE), and explored the precision effects in specific boundary areas. Ross and Stock (2019) applied Multilayer Perceptron (MLP) to explore the relationship between monthly marine elements and dissolved oxygen concentration in Chesapeake Bay, analyzing stratification phenomena on a sub-seasonal scale. However, the structure of machine learning is relatively simple, leaving considerable room for the further boosting of the fitting precision.

Recently, deep learning has been employed to increase the precision of MDOC inversion based on single point buoy data. Wang et al. (2020) used DJINN and its improved version, M-DJINN, to clarify the relationship between dissolved oxygen concentration and other variables such as temperature and salinity, utilizing data from the World Ocean Database. Experimental evidence shows that the precision of deep learning networks significantly outperforms traditional machine learning algorithms. However, the interpretability of deep learning networks is limited by their hidden layers, which extensively abstract and transform input data nonlinearly, and involve large number of parameter. This complexity makes it difficult to understand how the model operates. As a result, current high-precision MDOC inversion methods encounter difficulties in gaining full trust from decision-makers in marine ecology, posing substantial risks in decision-making processes. Therefore, developing a fully interpretable, high-precision intelligent inversion framework for MDOC becomes a significant challenge to overcome.

Rule-based methodologies offer a practical solution for achieving interpretable computations with high precision. Friedman and Popescu (2008) introduced RuleFit, a model consisting of a linear combination of rules and linear components, where each rule is expressed through straightforward evaluative statements about the input variables’ values. This collection of rules can achieve predictive precision comparable to the best methods, with the added benefit of being easily interpretable. In recent years, RuleFit has seen widespread use in fields that emphasize the interpretability of artificial intelligence, such as intelligent healthcare. For instance, Carrazana-Escalona et al. (2022) used RuleFit to predict the characteristics of blood pressure parameters among 8 adolescent volunteers during dynamic pressure-bearing processes, and Luo et al. (2022) applied it to diagnose nasopharyngeal carcinoma in 1706 patients. These studies illustrate that RuleFit can provide operational rules for models with high precision, thus enhancing their inherent interpretability. Although rule-based methods have the advantages of high precision and interpretability, these methods still have certain limitations. Bénard et al. (2021) introduced SIRUS, which enhances the precision and stability of rule extraction by restricting decision tree node splits to empirical quantile positions. However, the conclusions provided by this method are too specific and verbose, making it difficult to analyze the rules. Additionally, Mollas et al. (2022) proposed LionForests, which is valued for its “conclusiveness”, demonstrating improved stability and interpretability. Nonetheless, this method still has shortcomings in terms of its coverage of the decision-making process. Zhang et al. (2023) introduced OptExplain, an algorithm that utilizes particle swarm optimization for the optimization process, but it is currently applicable only to classification tasks, which does not align with the MDOC inversion task. By contrast, RuleFit, with its broad application base and superior performance, excels in rule generation and offers ease of interpretation. Therefore, RuleFit is ultimately selected as the inversion model in this work.

In this paper, we introduce a framework that offers both high precision and interpretability for the intelligent inversion of MDOC. We have named this framework CDRP because it comprises Causal Discovery, Drift Detection, RuleFit Model, and Post Hoc Analysis. To clarify the causal relationships between marine elements, we adopt the PCMCI causal discovery method, which helps to remove weakly correlated relationships and elucidate the associations between MDOC and other relevant elements, thereby enhancing the effectiveness and interpretability of model learning. In addition, the concept drift detection technique is also used to further improve the precision of the intelligent model and to help users understand the key features of the data. This technique helps users select more representative data, known as key frame data, for training the intelligent inversion model. Additionally, to realize high-precision interpretable intelligent inversion at the computational aspect of the model, we utilize the rule-based RuleFit algorithm. This algorithm not only achieves high-precision inversion of MDOC but also aids in clarifying the internal mechanisms of intelligent computation by analyzing the extracted rules. Upon completion of training, we utilize post-hoc analysis techniques like SHAP and LIME to investigate the model’s operational mechanisms. Our focus is on obtaining quantitative insights into how different marine elements influence the climatological normals of MDOC, both in terms of magnitude and direction. The analysis shows that our framework CDRP produces results that align well with conclusions in marine literature. In summary, our contributions are mainly in the following five aspects: (i) We propose an interpretable artificial intelligence framework CDRP for achieving high-precision interpretability in the MDOC intelligent inversion process; (ii) The introduction of PCMCI enhances the interpretability of CDRP by elucidating the relationships between MDOC and other elements while eliminating the interference of weakly correlated elements; (iii) By utilizing concept drift detection, this paper ensures a more representative selection of training data and model tuning, thereby effectively elevating the model’s precision and interpretability based on data reduction; (iv) This study pioneers the application of the rule-based RuleFit model to marine ecology, enhancing both the inversion precision and the interpretability of the operational mechanism; (v) The validation of CDRP through causal discovery, rule analysis, SHAP, and LIME, and its consistency with conclusions in marine literature on MDOC, effectively ensures interpretability in the inversion process.

2 Materials and methods

2.1 study area and dataset.

After reviewing extensive literature and datasets, we have preliminarily identified several marine elements related to MDOC inversion, including temperature, salinity, pH, chlorophyll concentration, turbidity, CO 2 concentration, water column level, and sediment phosphorus. In further selection of these elements, we have considered the following aspects: Since both pH and CO 2 concentration are key indicators of ocean acidification, which creates redundancy in their impact mechanisms on MDOC, we have abandoned the CO 2 concentration in our study. Additionally, as the oceanographic data involved in the inversion task are two-dimensional, while water column level are inherently three-dimensional, we will not consider the water column level for MDOC inversion. Moreover, because turbidity already reflects certain changes in sediment phosphorus, and data on sediment phosphorus are difficult to obtain, we have decided to exclude the utilization of sediment phosphorus. Finally, we selected the following 5 marine elements for further exploration: temperature (OTMP), salinity (SAL), chlorophyll concentration (CLCON), turbidity (TURB), and pH.

The dataset used in this study was provided by the National Data Buoy Center (NDBC) 3 , which is part of the National Oceanic and Atmospheric Administration (NOAA). It includes data collected from about 100 moored buoys and Coastal-Marine Automated Network (C-MAN) stations. Additionally, it includes data from 55 Tropical Atmosphere Ocean (TAO) buoys that are deployed and maintained in the equatorial Pacific, covering a range from 9°N to 8°S and from 95°W to 165°E. This buoy network system automatically captures and transmits real-time meteorological and oceanographic data to the National Ocean Service (NOS), located in Maryland.

2.2 Description of the proposed framework

In this research, we introduce a high-precision interpretable framework aimed at resolving the MDOC inversion challenge. This framework ( Figure 2 ) is principally segmented into four phases: Causal Discovery, Drift Detection, RuleFit Model, and Post Hoc Analysis. All interpretive actions are supported by validation from marine-related research literature, ensuring the professional integrity and logical consistency of the interpretive results. The details are outlined as follows:

Figure 2 Framework of the proposed approach.

Causal Discovery: In this stage, we utilize the causal discovery algorithm PCMCI to learn the causal relationship among the marine elements. By removing elements with weakly correlated relationship, we eliminate the interference with the model learning process. Subsequently, we analyze the associations between MDOC and relevant elements, enhancing the overall interpretability of the framework.

Drift Detection: This phase involves calculating the drift degree in continuously batched stream data to identify the timing (When) and specific data distribution (Where) of concept drift occurrences. It also includes an analysis of the underlying reasons related to marine observation processes (Why). The data collected when concept drift occurs are marked as key frame data for the intelligent model’s training.

RuleFit Model: Training the RuleFit model with the dataset refined by key frame selection in the previous phase, elevates the precision of inversion and allows for the extraction of the model’s internal operational rules. Analyzing these rules offers a preliminary explanation of the model’s operational mechanism.

Post Hoc Analysis: Employing global-level SHAP analysis and local-level LIME analysis offers more detailed explanations of the RuleFit model’s operational mechanism. The insights derived from causal discovery, RuleFit’s rules, along with these analyses, demonstrate excellent consistency with conclusions in marine literature, thus greatly enhancing the interpretability of intelligent computations.

2.3 Causal discovery

To elucidate the causal relationships between each marine element and MDOC, we introduced the PCMCI algorithm for causal discovery. This method was proposed by Runge et al. (2019) and consists of two main stages:

(i) PC Algorithm: Used for causal relationship discovery in time series data. It iteratively employs independence testing to remove unrelated causal associations, converging to a small number of key causal relationships and constructing an initial causal relationship graph.

(ii) MCI Algorithm: Used for instantaneous conditional independence testing. It suppresses false positives for highly interdependent time series.

Given a dynamic system X t = { X t 1 , … , X t N } of N representing marine elements considered at t time points, the following equation holds true ( Equation 1 ):

where f j represents some potential nonlinear functional dependencies, and n t j denotes mutually independent dynamic noise P ( X t j ) represents the causal parents of variable X t j among all N elements in the past. This causal discovery method is based on the concept of conditional independence. By estimating the strength and direction of causal relationships between highly interdependent time series of multiple marine elements, it effectively removes the interference of weakly correlated marine elements in model learning. Furthermore, by classifying each marine element based on its association with MDOC, the interpretability of the overall framework can be effectively enhanced.

2.4 Drift detection

The dissolved oxygen station data utilized in this paper is presented as a continuous data stream. As time progresses, the distribution of input data may undergo significant changes, which may adversely affect the performance of the intelligent inversion model trained on historical data. This phenomenon is known as concept drift ( Lu et al., 2018 ). Detecting concept drift enables adjustments to the intelligent inversion model to improve its precision. It also allows for explanations of changes in data distribution, linking these changes to variations in marine elements. The methodology employed in this paper utilizes incremental Gaussian Mixture Model (GMM) clustering for each data batch. This process calculates the drift degree between the current batch and historical marine data. It selects the most representative data exceeding a predefined threshold to compile a dataset, designated as key frame data, for training the inversion task ( Yang et al., 2020 ). The formula for drift degree is defined as follows ( Equation 2 ):

In this context, |·| represents the number of marine data samples, and d (·) denotes the energy distance between marine data samples. X t represents all the data of the current batch. X t i and X ^ i respectively represent the current batch data and the historical data for the i-th cluster. The formula for energy distance is defined as follows ( Equation 3 ):

Here, A = 1 m n ∑ i = 1 n ∑ j = 1 m ‖ x i − y i ‖ represents the average Euclidean distance between elements of marine features in two sets of marine data samples X and Y . B = 1 n 2 ∑ i = 1 n ∑ j = 1 n ‖ x i − x j ‖ and C = 1 m 2 ∑ i = 1 m ∑ j = 1 m ‖ y i − y j ‖ respectively represent the average Euclidean distance between elements within marine data samples X and Y .

2.5 Model computation

To elevate the computational precision and clarify the intrinsic operational mechanisms of the intelligent inversion model for MDOC, a rule-based ensemble method, RuleFit, has been adopted. This method constructs a model through a linear combination of rules and linear expressions, where each rule includes a concise set of statements about the individual input variables. Such collection of rules can achieve predictive precision comparable to that of the best methods. It also enables an initial understanding of the operational mechanism of the intelligent model through the analysis of principal rules ( Friedman and Popescu, 2008 ). Specifically, given a marine data sample x = { x 1 ,x 2 ,…,x n } T ∈ R n , the RuleFit model is defined as follows ( Wan et al., 2023 ) ( Equation 4 ):

Here, α 0 ∈   R ,   α k ∈   R ( k = 0 , 1 , … , K ) ,   α j ∗ ∈ R ( j = 1 , … , n ) represent the MDOC climatological normals, the coefficients for the rule terms of the intelligent inversion model, and the coefficients for the linear terms of intelligent inversion model, respectively. The rule terms are formed by combining judgment clauses for specific marine elements ( r k : R n → R ), and the linear terms are comprised of functions related to specific marine elements ( l j : R → R ).

2.6 Post hoc analysis

2.6.1 shap analysis.

To provide a more comprehensive and reliable explanation of the operational mechanisms of the inversion model, we utilized SHAP (Shapley Additive Explanations) analysis. This approach quantitatively assesses the impact magnitude and direction that various marine elements have on the MDOC climatological normals from a global perspective. SHAP represents a game theory-based method for interpreting artificial intelligence models ( Štrumbelj and Kononenko, 2014 ). It facilitates assessing the negative and positive effects that marine elements have on the output of the intelligent MDOC inversion model. Given an intelligent inversion model trained with marine data samples X i = { x 1 ,x 2 ,…,x n } T , an explanation model (EM) is employed by SHAP to evaluate the contribution of each marine element to the intelligent inversion model. The details can be described in the following equation ( Equations 5 , 6 ):

Where n is the number of marine elements, t i is the simplification of marine element i , t i ∈ R denotes the contribution of variable i to the artificial intelligence model, \ denotes the difference-set notation for set operations, and f indicates the interpretable artificial intelligence model.

2.6.2 LIME analysis

To better understand how intelligent inversion models work, especially at critical points like MDOC extrema, we intend to utilize the Local Interpretable Model-agnostic Explanations (LIME) analysis. This approach will allow us to examine how various marine elements influence the output of the intelligent inversion model. When a new observation is introduced, LIME creates an extended dataset consisting of perturbed samples and their corresponding model outputs. A linear explanatory model is then adjusted based on this dataset, applying weights according to the closeness of these sampled observations ( Ribeiro et al., 2016 ). Through this approach, we can apply the interpretable model, which is tailored for local explanations ( Chakraborty et al., 2021 ), to estimate the influence of marine elements on the MDOC extrema. Specifically, the definition of the local interpretable model g is as follows ( Equation 7 ):

Here, π x measures how close the changed marine data instances are to each other, usually using a Gaussian kernel. L ( f,g,π x ) shows how much the interpretable model g differs from the model f we want to explain, especially at MDOC extrema. Ω( g ) measures the complexity of the interpretable model (such as the number of non-zero weights in a linear model).

2.7 Implementation and evaluation metrics

In this study, we applied the Python programming language, widely used in data science, along with key modules such as Numpy, scikit-learn, SHAP, and LIME. The configuration of our environment comprised Python 3.7, a 12th Gen Intel(R) Core(TM) i7–12700H, and Windows 11.

To evaluate the precision of CDRP within the MDOC inversion task, this study utilizes three widely recognized statistical and regression metrics to measure its performance: Mean Square Error (MSE), Accuracy (ACC), and Explained Variance Score (EVS).

Mean Squared Error (MSE) is defined as follows ( Equation 8 ):

In this context, n denotes the total number of observations, with y i indicating the observed value for the i-th observation, and y ^ i representing the predicted value for it. A reduction in the MSE value signifies improved accuracy in the inversions. Consequently, the accuracy of the model is derived from the MSE as defined below ( Equation 9 ):

Explained variance score (EVS) is defined as follows ( Equation 10 ):

Here, Y ^ denotes the predicted output, Y is the observed output in relation to Y ^ , and Var represents variance. Importantly, the highest possible score is 1, with a lower score reflecting a decrease in the prediction’s adequacy, as shown by the variance in the dependent variables.

3.1 Inversion performance

In this study, buoy data from various locations along the U.S. West Coast were employed, with the dataset for MDOC inversion comprising the first-hour average values of temperature (OTMP), salinity (SAL), chlorophyll concentration (CLCON), and pH (According to the causal discovery graph in Section 3.2.1, we excluded turbidity, which showed weak correlation with MDOC). SVR, RFR, MLP, DJINN, M-DJINN, and RuleFit were trained and evaluated utilizing this dataset. To validate the models’ performance in this study, data from January 1, 2016, to October 18, 2019, making up the initial continuous 80%, was designated as the training set. Conversely, data ranging from October 19, 2019, to February 22, 2022, representing the subsequent continuous 20%, was selected for the test set.

By continuously conducting concept drift detection on marine element data, organized in batches corresponding to one week’s duration, variations in drift degree are depicted in Figure 3 . By setting an appropriate threshold, it becomes possible to accurately identify the dates when drifts occur. Following the identification of concept drift, data from those dates are merged with the dataset previously used for model training. The model is then retrained on this updated dataset. Upon setting the drift degree threshold at 12, eight specific instances of concept drift were detected. Figure 4 shows the distinct changes in data distribution for each instance of concept drift, highlighted by red lines. Based on comparisons of data before and after each detected instance of concept drift, preliminary analysis of the causative factors is presented as follows: The first concept drift occurred on March 12, 2016 ( Figure 4A ), where the pH decreased from 8.6 to 8.0. In contrast, on January 24, 2019 ( Figure 4G ), pH showed an increase. These changes may be attributed to the influence of upwelling and possible coastal discharge ( Kroeker et al., 2020 ; Li et al., 2023a ). On June 10, 2016 ( Figure 4B ), October 9, 2016 ( Figure 4D ), and March 28, 2019 ( Figure 4H ), there were significant increases in chlorophyll concentration, possibly due to the extensive proliferation of phytoplankton ( Conley et al., 2007 ). Furthermore, fluctuations in temperature and salinity influenced by local coastal climate were effectively detected and corrected ( Figures 4C, E, F ).

Figure 3 The variation of drift degree over time. (The red line represents the set drift threshold, and the red solid point is the dates when the concept drift detection is occurred).

Figure 4 (A–H) The data distribution at the date of concept drift. (The date when concept drift occurs is highlighted by a red line).

This research selected MDOC data samples from the first 30 days to pre train the inversion model, and then conducted concept drift detection in batches of 7 days. As shown in Figure 3 , eight distinct instances of concept drift were detected. Together with 30 pre-trained data samples, the final key frame dataset comprised a total of 86 data samples, which shows a significant reduction compared to the overall 1080 MDOC training data samples. It is worth noting that the data reduction reduced the complexity and redundancy of training data, filtered out more informative features, and allowed the model to focus more on learning key features, effectively helping users improve their understanding of the overall features of the learning process ( Atitey et al., 2024 ). Therefore, while improving the precision of MDOC inversion, it effectively enhanced the interpretability of the intelligent inversion process.

To assess the precision superiority of CDRP for MDOC inversion task, a comparative analysis was conducted between CDRP and several models currently used in this field, including SVR, RFR, MLP, DJINN, and M-DJINN. Despite the moderate novelty of the models compared in the experiment, they adequately represent the current accuracy level in the field of MDOC inversion. Therefore, the experimental results can convincingly demonstrate the superior precision of CDRP. Considering the diversity of hyperparameters among the models used in our experiment, we adopted the hyperparameter settings recommended in their respective studies. The hyperparameter configurations are detailed in Table 1 . It can be observed that the hyperparameters for machine learning algorithms are relatively simpler, whereas those for deep learning methodologies are more complex. This preliminary observation reflects that deep learning needs a large amount of data and parameters, leading to higher precision but lower interpretability ( Li et al., 2024b ).

Table 1 Hyperparameter configurations of the employed models.

To explore the effect of different training strategies on the precision of MDOC inversion, we trained selected models using four distinct strategies: direct training, training with causal discovery, training with concept drift detection, and training with a combination of causal discovery and concept drift detection. The experimental results are presented in Table 2 . Our analysis reveals that integrating causal discovery significantly improved the inversion precision across all participating models, achieving optimal performance in most cases. This highlights the effect of removing weak-correlation factors on enhancing precision, with implementation of causal discovery described in Section 3.2.1. It is speculated that this is due to the weak correlation between turbidity and MDOC, as well as its characteristics of large variability and unstable changes, which may cause negative interference to the inversion model ( Schmitt et al., 2008 ). Besides, introducing concept drift detection notably benefited the precision of tree-based algorithms (RuleFit, DJINN, and RFR). Especially in the training of the RFR model, incorporating concept drift detection achieved optimal accuracy. This occurred because when the training and test sets cover different time periods, tree-based algorithms can better learn generalizable mappings from more representative training data, which improves their performance on future tasks. Finally, we attempted to train the model using the strategy of concept drift detection update after removing the weakly associated turbidity. After analyzing the experimental results, it was found that not all models experienced further improvements in precision. This may be because the combination of causal discovery and concept drift detection update training strategies does not work well for all algorithms. Ultimately, it can be found that CDRP, which integrates causal discovery and concept drift detection within the RuleFit model, achieved the highest precision among the implemented methods.

Table 2 Performance of the selected models with different training strategies.

To validate the effectiveness of removing interference from weakly associated elements and concept drift detection on the improvement of RuleFit’s precision in the MDOC inversion tasks, RuleFit was trained using both direct training and training with a combination of causal discovery and concept drift detection. Figure 5 shows the variation curves of MSE, ACC, and EVS. It’s evident that the utilization of causal discovery and concept drift detection notably reduced MSE, while at the same time increasing ACC and EVS. Models enhanced with causal discovery and concept drift detection demonstrated significant early-stage optimization in MSE and ACC around 15 to 18 weeks, compared to the RuleFit model that underwent direct training. The EVS also showed an increase after the final training session was completed. This convincingly confirms the superiority of CDRP in elevating the precision for MDOC inversion task.

Figure 5 Performance of CDRP and directly trained RuleFit. (A) MSE; (B) ACC; (C) EVS. (The blue line represents the metric change of CDRP, while the solid red line represents the metric of directly trained RuleFit.).

For an intuitive analysis of CDRP’s fitting effect, the RuleFit model, trained with a combination of causal discovery and concept drift detection, was used to process the entire dataset. The comparison between predicted and observed MDOC values is presented through overlay plot and scatter plot ( Figure 6 ). Figure 6A displays a notable consistency between predicted and observed MDOC values. Meanwhile, Figure 6B reveals that the predictions for a significant portion of data points lie within the orange area, which signifies the range of Root Mean Square Error (RMSE). From the analysis, it can be concluded that CDRP demonstrates commendable fitting efficacy, rendering it applicable for real-world MDOC inversion task.

Figure 6 Comparison of the predicted and observed MDOC, using overlay plot (A) and scatter plot (B) . (Orange line in scatter plot is the fitted linear between observed and predicted values).

3.2 Interpretation of inversion results

3.2.1 causal discovery.

To analyze the correlation between marine elements from the causal perspective, we employ the PCMCI causal discovery algorithm to conduct causal analysis on the initially selected five elements (OTMP, SAL, CLCON, PH, and TURB) with the target element MDOC. Figure 7A shows the causal relationship graph between marine elements, while Figure 7B displays the causal relationship graph from the perspective of time series, highlighting the two-day delay between temperature (OTMP) and MDOC. In addition to the direct causal influence from pH to MDOC, chlorophyll concentration (CLCON) indirectly affects MDOC through OTMP. All of them indicate that pH, OTMP and CLCON are key elements in MDOC inversion. Additionally, a notable causal link exists from MDOC to salinity (SAL), which is supported by the subsequent SHAP analysis. Finally, we can conclude that turbidity (TURB) has almost no causal relationship with MDOC. This conclusion is reinforced by the experiments described in Section 3.1, which shows that removing weakly associated turbidity effectively reduces interference in the MDOC inversion task.

Figure 7 Causal relationships between marine elements. (A) Causal relationship graph; (B) Time series graph.

3.2.2 Rulefit rule extraction

RuleFit, composed of a series of readily interpretable IF-THEN rules and linear adjustments, not only demonstrates significant precision in inversion tasks but also enables initial insights into the operational mechanisms of the intelligent inversion. This is achieved through the extraction and subsequent analysis of critically significant rules. The primary rules extracted by RuleFit are detailed in Table 3 . By analyzing the judgments on specific elements within the primary rules, we conclude the following insights: (i) An increase in temperature is inversely related to MDOC, as shown by rules 3 and 4; (ii) Salinity mostly has a negative impact on MDOC, as depicted by rules 1, 2 and 3; (iii) A decrease in pH is related with a reduction in MDOC, as specified by rule 1 and 4; (iv) An elevation in chlorophyll concentration is positively linked to MDOC, as illustrated by rules 2.

Table 3 The main rules extracted through RuleFit.

3.2.3 SHAP analysis

To understand the operational mechanisms of the intelligent inversion model from a global perspective, SHAP analysis was utilized to quantify the impact magnitude and direction of the marine elements on the climatological normals of MDOC. We present a summary plot of the SHAP analysis for selected marine elements ( Figure 8 ). In this plot, the vertical axis orders the marine elements by their impact magnitude, and the horizontal axis shows the change (Shapley value) to the MDOC climatological normals (7.98 mg/L) based on the values of these marine elements. The color of the dots is detailed in the legend to the right of the plot, while the vertical stacking of dots illustrates the frequency of sample points with specific values. Figure 8 shows that salinity has the largest impact on MDOC climatological normals, with a trend suggesting that lower salinity leads to a higher positive impact on MDOC climatological normals. This is consistent with the results of RuleFit rule extraction analysis. The influence of pH on MDOC climatological normals is secondary, primarily indicating a positive impact at higher pH levels. Lower temperatures are associated with a greater positive impact on the MDOC climatological normals. The impact of chlorophyll concentration on MDOC climatological normals is the smallest, primarily manifested as a negative effect.

Figure 8 Summary plot for marine elements.

To investigate the interaction between salinity, which contributes most significantly to the impact on MDOC climatological normals, and other elements, we conducted SHAP analysis and produced dependence plots ( Figure 9 ). Preliminary analysis of Figure 9 allows us to deduce that salinity exerts a positive impact on the MDOC climatological normals when below approximately 25 psu, and manifests a negative impact when exceeding this threshold. The increase in temperature reduces the positive impact of salinity ( Figure 9A ), while an increase in pH elevates the positive impact of salinity ( Figure 9B ). Conversely, chlorophyll concentration does not exhibit a significant effect on the impact of salinity ( Figures 9C ).

Figure 9 Dependence plot of the interaction between SAL and other marine elements. (A) OTMP; (B) pH; (C) CLCON.

3.2.4 LIME analysis

Contrary to the SHAP analysis method, which focuses on assessing the global contributions of marine elements, LIME analysis can provide local interpretation for the influencing factors of various marine elements at key climate nodes, and the critical value range can guide the quantitative judgment of the impact direction of input elements on the output target element — MDOC, thereby providing local interpretation schemes and enhancing the interpretability of the overall framework. We selected three consecutive dates of MDOC minima ( Figures 10A-C ) and three consecutive dates of MDOC maxima ( Figures 10D-F ) to analyze the impact direction and magnitude of each marine elements within their respective value ranges at these critical climate nodes. We referenced the research by El Bilali et al. (2023) in our analysis, identifying the critical value ranges at which the direction of the influence of marine elements on the MDOC climatological normals changes. The analysis provides the following insights: (i) Salinity has the most significant impact on MDOC climatological normals, followed by pH in typical cases, with temperature being less significant, and chlorophyll concentration having the least impact; (ii) Salinity below 23.81 psu has a positive impact on the MDOC climatological normals, while levels above 29.70 psu have a negative impact; (iii) Temperatures below 12.64°C have a positive effect on the MDOC climatological normals, whereas temperatures above it have a negative effect; (iv) pH above 7.80 positively impacts the MDOC climatological normals, while pH below it has a negative impact; (v) Chlorophyll concentrations above 5.50 µ g/L positively affect the MDOC climatological normals, whereas in other circumstances a negative impact occurs.

Figure 10 (A–F) The impact of marine elements at the extrema of MDOC. Negative LIME values indicate MDOC below historical median while positive LIME values indicate MDOC above historical median.

4 Discussion

MDOC is one of the primary indicators in the domain of marine ecology. In this study, we applied a series of artificial intelligence models to the MDOC inversion task, where our proposed framework CDRP demonstrated optimal precision. SVR is sensitive to the characteristics of input data, while RFR is susceptible to overfitting. Moreover, complex models such as MLP, DJINN, and its modified version M-DJINN require large volumes of data for effective training. Conversely, the RuleFit model creates a broad set of predictive rules tailored for the inversion task. This method offers deep insights into the computational mechanisms of inversion and exhibits strong generalization abilities, as evidenced in ( Luo et al., 2022 ). Overall, CDRP which utilizes the RuleFit model achieves superior precision in inversion task.

4.1 Precision influenced by causal discovery and concept drift

We introduced causal discovery and concept drift detection during the training process. By analyzing the experimental results ( Table 2 ), it can be found that causal discovery significantly improves the inversion performance of all models. It follows that causal discovery can keenly identify uncorrelated elements, which can guide the improvement of model training strategies. However, concept drift detection only achieves a boost in effectiveness in tree-based algorithms, which demonstrates a kind of compatibility between them. By expanding to other tree-based algorithms, it is still possible to improve model performance while extracting key features of the dataset. Finally, CDRP achieves optimal performance by simultaneously introducing causal discovery and concept drift detection in training process. However it may be not the optimal training strategy for all models. Causal discovery involves input feature-level reduction from the perspective of causal inference, while concept drift detection involves time series-level reduction from the perspective of data distribution changes. This could potentially lead to excessive reduction and result in model underfitting.

4.2 Interpretability

In the literature on applying AI models to MDOC inversion task, there is a lack of exploration on interpretability. This can lead to risks associated with unknown computational logic. Therefore, enhancing the interpretability of the MDOC inversion task is of significant importance. To solve this problem, we constructed an interpretive process of “causal discovery + rules analysis + post hoc analysis + literature validation of consistency” to comprehensively improve the interpretability of the MDOC inversion process.

4.2.1 Causal discovery

To investigate the causal relationships between each marine element and MDOC, we introduced PCMCI to conduct causal discovery on temperature, salinity, pH, chlorophyll concentration, turbidity, and MDOC. By estimating the strength and directionality of causal relationships among highly interdependent time series of multiple marine elements, we found that PCMCI can effectively eliminate the interference of weakly correlated marine elements on model learning. After causal analysis, the causal graph and time series causal graph are shown in Figure 7 . By analyzing the correlation with MDOC, we can classify the marine elements into four categories: direct causal association (temperature, pH), indirect causal association (chlorophyll concentration), correlated association (salinity), and weakly correlated association (turbidity). Among these, temperature belongs to the category of time lagged causal correlation. Specifically, the temperature from two days ago has a direct causal effect on the current MDOC. The global warming and ocean acidification are direct factors leading to the occurrence of marine hypoxia, which is consistent with the results of causal graph analysis ( Breitburg et al., 2018 ; George et al., 2024 ). The promoting effect of chlorophyll on MDOC is essentially achieved through biomass influencing temperature, thus resulting in a positive correlation effect on MDOC ( MacPherson et al., 2007 ; Li et al., 2024a ). Although salinity is not a causal parent of MDOC, the causal relationship between MDOC and salinity makes this correlation relationship indispensable in the MDOC inversion task. Furthermore, the subsequent SHAP analysis further confirms the importance of salinity. Ultimately, turbidity does not exhibit significant causal relationship with other marine elements. Therefore, the presence of this element would bring negative interference to the MDOC inversion task. The quantitative experiments conducted earlier demonstrate that removing the interference from the turbidity significantly contributes to improving the accuracy level of MDOC inversion. This further elucidates the importance and necessity of introducing causal discovery method for enhancing interpretability.

4.2.2 Rules analysis

To analyze the influence of marine elements on the MDOC climatological normals from the model inference perspective, the RuleFit model was introduced. By establishing a large initial set of predictive rules and then refining these rules to improve inversion precision, this method achieves high-precision and also helps understand how the model works. This understanding comes from utilizing and interpreting the set of rules. The decrease in MDOC with higher temperature is due to increased oxygen demand and reduced oxygen solubility as temperature rise ( Breitburg et al., 2018 ; Ye et al., 2021 , 2023 ; Bandara et al., 2024 ). Silva et al. (2009) described Equatorial Subsurface Water (ESSW) characteristics, noting that the highest underwater salinity values are associated with the lowest MDOC and high nitrate and phosphate levels. This supports the idea that colder, less saline water can dissolve more ( Kouketsu et al., 2022 ; Sun et al., 2023 ), which matches the RuleFit rule showing an inverse relationship between salinity and MDOC. The research by Schmitt et al. (2008) shows long-range correlations between pH and MDOC in their power-law spectrum, particularly noting that ocean acidification goes along with marine hypoxia ( Gao et al., 2020 ; George et al., 2024 ). This supports the RuleFit finding that a lower pH results in decreased MDOC. The rule that an increase in chlorophyll concentration leads to higher MDOC is supported by ( MacPherson et al., 2007 ; Li et al., 2024a ), indicating that higher chlorophyll concentration produce more oxygen indirectly, thereby increasing MDOC. Therefore, it can be concluded that the RuleFit model utilized in CDRP extracts rules that are easily interpretable with high precision, and this approach is well-supported by a wealth of marine scientific literature.

4.2.3 Post-hoc analysis

SHAP analysis is employed to enhance the understanding of the MDOC inversion process by examining its results. This analysis, conducted from a global perspective, explores how marine elements contribute to the MDOC inversion task. Additionally, a local interpretability analysis through LIME is utilized to analyze the model’s computational basis at MDOC extrema. SHAP and LIME analyses show that marine elements can affect MDOC climatological normals positively or negatively at different times. Seasonal changes in MDOC are influenced by sunlight, ice cover, air temperature, winds, and currents ( Kroeker et al., 2020 ; Xu et al., 2022 ). Events like upwelling, which brings colder deep seawater with lower MDOC content to the surface, also cause short-term MDOC variations ( Booth et al., 2012 ; Chen et al., 2022 ; Castrillón-Cifuentes et al., 2023 ; Wang et al., 2023a ). The conclusion drawn from the SHAP analysis that salinity is negatively correlated with MDOC is consistent with RuleFit analysis. Furthermore, the findings about temperature’s negative impact and pH’s positive impact on salinity contributions agree with previous analysis based on RuleFit rules. Moreover, LIME analysis identifies the critical value range for salinity’s impacts as 23.81–29.70 psu. This range includes zero Shapley value of salinity from SHAP analysis (around 25 psu, as shown in Figure 9 ), confirming the consistency between SHAP and LIME analysis on salinity. Similarly, LIME provides the direction of impact (positive, negative, positive) and critical value ranges for pH, temperature and chlorophyll concentration on the MDOC climatological normals (7.80, 12.64°C, 5.50 µ g/L), respectively.

4.2.4 Insight from interpretability analysis

Post-hoc analysis has led to findings that align with causal discovery and RuleFit rules, and they also offer specific critical value ranges. These findings provide marine scientists with quantitative insights into how various marine elements influence the magnitude and direction of changes in MDOC climatological normals. Based on insights from interpretability analysis, we can propose several strategies to reduce marine hypoxia. The finding that salinity negatively impacts MDOC indicates that reducing sewage discharge could help prevent deoxygenation in the ocean, especially near coastlines. Similarly, limiting greenhouse gas emissions to slow down global warming and ocean acidification is also an effective strategy. Moreover, maintaining marine ecological indicators within reasonable ranges is crucial for controlling dissolved oxygen levels.

5 Conclusion

This paper introduces an interpretable artificial intelligence framework CDRP designed for high-precision MDOC inversion. Initially, PCMCI is utilized for causal discovery of marine elements and to eliminate the interference of weakly associated elements. Following that, key frame data is selected through concept drift detection, resulting in the formation of the training dataset. Subsequently, the dataset is fed into the rule-based RuleFit model for training. This step is followed by extracting operational rules, which enables the establishment of an initial interpretation. Afterwards, an advanced analysis is conducted utilizing post-hoc analysis techniques, specifically SHAP and LIME. This comprehensive approach offers insights that are consistent with actual marine observation, especially in terms of their influence on the MDOC climatological normals. In comparative tests with SVR, RFR, MLP, DJINN, and M-DJINN, our framework showed the best performance in precision and interpretability. The principal findings from the analysis of research results are as follows: (i) Conducting causal discovery of marine elements through PCMCI, along with removing weakly associated elements and analyzing causal relationships, can effectively enhance the effectiveness and interpretability of model learning. (ii) Using concept drift detection to capture changes in marine elements effectively enhances the precision and interpretability based on data reduction of CDRP. (iii) Considering RuleFit, SHAP, and LIME analysis results together, the ranking of the influence of marine elements on MDOC climatological normals is: salinity > pH > temperature > chlorophyll concentration. (iv) The critical value ranges for the impact on climatological normals are salinity (23.81-29.70 psu), pH (7.80), temperature (12.64°C) and chlorophyll concentration (5.50 µ g/L). In summary, CDRP demonstrates high precision and interpretability in single-point measured MDOC inversion tasks, displaying commendable consistency with conclusions in marine literature on MDOC.

Currently, considering the expansion of remote sensing data sources, exploring computational techniques that improve both precision and interpretability with this data is seen as a promising field for academic research. Furthermore, a more valuable interpretation of concept drift phenomena can be achieved through deep involvement in causal analysis. Therefore, conducting further analysis at the moment when concept drift occurs through methods such as causal discovery and causal effect analysis represents a highly prospective research direction.

Data availability statement

Publicly available datasets were analyzed in this study. This data can be found here: https://www.ndbc.noaa.gov/historical_data.shtml .

Author contributions

XL: Writing – original draft, Writing – review & editing, Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Validation, Visualization. ZL: Writing – original draft, Writing – review & editing, Data curation, Investigation, Software, Validation. ZY: Writing – review & editing, Software, Supervision, Validation. FM: Writing – review & editing, Methodology, Supervision, Validation. TS: Writing – review & editing, Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation.

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by grants of National Key Research and Development Project of China (Project No. 2021YFA1000103), the Natural Science Foundation of Shandong Province of China (Project No. ZR2020MF140) and the Key Laboratory of Marine Hazard Forecasting, Ministry of Natural Resources (Project No. LOMF2202).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

• ^ https://archimer.ifremer.fr/doc/00187/29825/
• ^ https://www.ncei.noaa.gov/products/world-ocean-database
• ^ https://www.ndbc.noaa.gov/historicaldata.shtml

Atitey K., Motsinger-Reif A. A., Anchang B. (2024). Model-based evaluation of spatiotemporal data reduction methods with unknown ground truth through optimal visualization and interpretability metrics. Briefings Bioinf. 25, bbad455. doi: 10.1093/bib/bbad455

CrossRef Full Text | Google Scholar

Bandara R. M. W. J., Curchitser E., Pinsky M. L. (2024). The importance of oxygen for explaining rapid shifts in a marine fish. Global Change Biol. 30, e17008. doi: 10.1111/gcb.17008

Bénard C., Biau G., Da Veiga S., Scornet E. (2021). “Interpretable random forests via rule extraction,” in International Conference on Artificial Intelligence and Statistics (PMLR). Microtome Publishing31: USA. 937–945.

Google Scholar

Booth J. A. T., McPhee-Shaw E. E., Chua P., Kingsley E., Denny M., Phillips R., et al. (2012). Natural intrusions of hypoxic, low ph water into nearshore marine environments on the california coast. Continent. Shelf Res. 45, 108–115. doi: 10.1016/j.csr.2012.06.009

Breitburg D., Levin L. A., Oschlies A., Grégoire M., Chavez F. P., Conley D. J., et al. (2018). Declining oxygen in the global ocean and coastal waters. Science 359, eaam7240. doi: 10.1126/science.aam7240

PubMed Abstract | CrossRef Full Text | Google Scholar

Brock A. L., Kostadinova K., Mørk-Pedersen E., Hensel F., Zhang Y., Valverde-Pérez B., et al. (2023). Remediation of marine dead zones by enhancing microbial sulfide oxidation using electrodes. Mar. pollut. Bull. 193, 115142. doi: 10.1016/j.marpolbul.2023.115142

Carrazana-Escalona R., Andreu-Heredia A., Moreno-Padilla M., Reyes del Paso G. A., Sánchez- Hechavarría M. E., Muñoz-Bustos G. (2022). Blood pressure prediction using ensemble rules during isometric sustained weight test. J. Cardiovasc. Dev. Dis. 9, 440. doi: 10.3390/jcdd9120440

Castrillón-Cifuentes A. L., Zapata F. A., Wild C. (2023). Physiological responses of pocillopora corals to upwelling events in the eastern tropical pacific. Front. Mar. Sci. 10. doi: 10.3389/fmars.2023.1212717

Chakraborty D., Başağaoğlu H., Winterle J. (2021). Interpretable vs. noninterpretable machine learning models for data-driven hydro-climatological process modeling. Expert Syst. Appl. 170, 114498. doi: 10.1016/j.eswa.2020.114498

Chen C.-C., Ko D. S., Gong G.-C., Lien C.-C., Chou W.-C., Lee H.-J., et al. (2022). Reoxygenation of the hypoxia in the east China sea: A ventilation opening for marine life. Front. Mar. Sci. 8. doi: 10.3389/fmars.2021.787808

Conley D. J., Carstensen J., Ærtebjerg G., Christensen P. B., Dalsgaard T., Hansen J. L., et al. (2007). Long-term changes and impacts of hypoxia in danish coastal waters. Ecol. Appl. 17, S165–S184. doi: 10.1890/05-0766.1

El Bilali A., Abdeslam T., Ayoub N., Lamane H., Ezzaouini M. A., Elbeltagi A. (2023). An interpretable machine learning approach based on dnn, svr, extra tree, and xgboost models for predicting daily pan evaporation. J. Environ. Manage. 327, 116890. doi: 10.1016/j.jenvman.2022.116890

Friedman J. H., Popescu B. E. (2008). Predictive learning via rule ensembles. Ann. Appl. Stat 2, 916–954. doi: 10.1214/07-AOAS148

Gao K., Gao G., Wang Y., Dupont S. (2020). Impacts of ocean acidification under multiple stressors on typical organisms and ecological processes. Mar. Life Sci. Technol. 2, 279–291. doi: 10.1007/s42995-020-00048-w

George R., Job S., Semba M., Monga E., Lugendo B., Tuda A. O., et al. (2024). High-frequency dynamics of ph, dissolved oxygen, and temperature in the coastal ecosystems of the tanga-pemba seascape: implications for upwelling-enhanced ocean acidification and deoxygenation. Front. Mar. Sci. 10. doi: 10.3389/fmars.2023.1286870

Giglio D., Lyubchich V., Mazloff M. R. (2018). Estimating oxygen in the southern ocean using argo temperature and salinity. J. Geophys. Res.: Oceans 123, 4280–4297. doi: 10.1029/2017JC013404

Hutchins D. A., Capone D. G. (2022). The marine nitrogen cycle: new developments and global change. Nat. Rev. Microbiol. 20, 401–414. doi: 10.1038/s41579-022-00687-z

Ji X., Shang X., Dahlgren R. A., Zhang M. (2017). Prediction of dissolved oxygen concentration in hypoxic river systems using support vector machine: a case study of Een-Rui Tang river, China. Environ. Sci. pollut. Res. 24, 16062–16076. doi: 10.1007/s11356-017-9243-7

Jiang Y., Gou Y., Zhang T., Wang K., Hu C. (2017). A machine learning approach to argo data analysis in a thermocline. Sensors 17, 2225. doi: 10.3390/s17102225

Jin X., Gruber N. (2003). Offsetting the radiative benefit of ocean iron fertilization by enhancing n2o emissions. Geophys. Res. Lett. 30. doi: 10.1029/2003GL018458

Johnson K. S., Plant J. N., Coletti L. J., Jannasch H. W., Sakamoto C. M., Riser S. C., et al. (2017). Biogeochemical sensor performance in the soccom profiling float array. J. Geophys. Res.: Oceans 122, 6416–6436. doi: 10.1002/2017JC012838

Karadurmus U., Sari M. (2022). Marine mucilage in the sea of Marmara and its effects on the marine ecosystem: mass deaths. Turkish J. Zool. 46, 93–102. doi: 10.3906/zoo-2108–14

Kouketsu S., Murata A., Arulananthan K. (2022). Subsurface water property structures along 80 e under the positive Indian ocean dipole mode in december 2019. Front. Mar. Sci. 9. doi: 10.3389/fmars.2022.848756

Kroeker K. J., Bell L. E., Donham E. M., Hoshijima U., Lummis S., Toy J. A., et al. (2020). Ecological change in dynamic environments: Accounting for temporal environmental variability in studies of ocean change biology. Global Change Biol. 26, 54–67. doi: 10.1111/gcb.14868

Li H., Li X., Song D., Nie J., Liang S. (2024a). Prediction on daily spatial distribution of chlorophylla in coastal seas using a synthetic method of remote sensing, machine learning and numerical modeling. Sci. Total Environ. 910, 168642. doi: 10.1016/j.scitotenv.2023.168642

Li Y., Robinson S. V., Nguyen L. H., Liu J. (2023b). Satellite prediction of coastal hypoxia in the northern gulf of Mexico. Remote Sens. Environ. 284, 113346. doi: 10.1016/j.rse.2022.113346

Li X., Wang Z., Chen C., Tao C., Qiu Y., Liu J., et al. (2024b). Semid: Blind image inpainting with semantic inconsistency detection. Tsinghua Sci. Technol. 29, 1053–1068. doi: 10.26599/TST.2023.9010079

Li X., Wang F., Song T., Meng F., Zhao X. (2023a). Astmen: an adaptive spatiotemporal and multi-element fusion network for ocean surface currents forecasting. Front. Mar. Sci . 10. doi: 10.3389/fmars.2023.1281387

Lu J., Liu A., Dong F., Gu F., Gama J., Zhang G. (2018). Learning under concept drift: A review. IEEE Trans. knowledge Data Eng. 31, 2346–2363. doi: 10.1109/TKDE.2018.2876857

Luo C., Li S., Zhao Q., Ou Q., Huang W., Ruan G., et al. (2022). Rulefit-based nomogram using inflammatory indicators for predicting survival in nasopharyngeal carcinoma, a bi-center study. J. Inflammation Res. 15, 4803–4815. doi: 10.2147/JIR.S366922

MacPherson T. A., Cahoon L. B., Mallin M. A. (2007). Water column oxygen demand and sediment oxygen flux: patterns of oxygen depletion in tidal creeks. Hydrobiologia 586, 235–248. doi: 10.1007/s10750-007-0643-4

Matear R., Hirst A. (2003). Long-term changes in dissolved oxygen concentrations in the ocean caused by protracted global warming. Global Biogeochem. Cycles 17. doi: 10.1029/2002GB001997

Mollas I., Bassiliades N., Tsoumakas G. (2022). Conclusive local interpretation rules for random forests. Data Min. Knowledge Discovery 36, 1521–1574. doi: 10.1007/s10618-022-00839-y

Naik V., Manjappa S. (2011). Mathematical modeling for dissolved oxygen prediction in mixing zone–a case study. J. Environ. Sci. Eng. 53, 429–436.

PubMed Abstract | Google Scholar

Ribeiro M. T., Singh S., Guestrin C. (2016). ““ why should i trust you?” explaining the predictions of any classifier,” in Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. ASSOC Computing Machinery 1515: NEW YORK USA. 1135–1144. doi: 10.1145/2939672.2939778

Ross A. C., Stock C. A. (2019). An assessment of the predictability of column minimum dissolved oxygen concentrations in chesapeake bay using a machine learning model. Estuarine Coast. Shelf Sci. 221, 53–65. doi: 10.1016/j.ecss.2019.03.007

Runge J., Nowack P., Kretschmer M., Flaxman S., Sejdinovic D. (2019). Detecting and quantifying causal associations in large nonlinear time series datasets. Sci. Adv. 5, eaau4996. doi: 10.1126/sciadv.aau4996

Schmitt F. G., Dur G., Souissi S., Zongo S. B. (2008). Statistical properties of turbidity, oxygen and ph fluctuations in the seine river estuary (France). Phys. A: Stat. Mech. its Appl. 387, 6613–6623. doi: 10.1016/j.physa.2008.08.026

Silva N., Rojas N., Fedele A. (2009). Water masses in the humboldt current system: Properties, distribution, and the nitrate deficit as a chemical water mass tracer for equatorial subsurface water off Chile. Deep Sea Res. Part II: Topical Stud. Oceanogr. 56, 1004–1020.

Štrumbelj E., Kononenko I. (2014). Explaining prediction models and individual predictions with feature contributions. Knowledge Inf. Syst. 41, 647–665. doi: 10.1007/s10115-013-0679-x

Sun Y.-J., Robinson L. F., Parkinson I. J., Stewart J. A., Lu W., Hardisty D. S., et al. (2023). Iodine-to-calcium ratios in deep-sea scleractinian and bamboo corals. Front. Mar. Sci. 10. doi: 10.3389/fmars.2023.1264380

Suntharalingam P., Sarmiento J. L., Toggweiler J. (2000). Global significance of nitrous-oxide production and transport from oceanic low-oxygen zones: A modeling study. Global Biogeochem. Cycles 14, 1353–1370. doi: 10.1029/1999GB900100

Wan K., Tanioka K., Shimokawa T. (2023). Rule ensemble method with adaptive group lasso for heterogeneous treatment effect estimation. Stat Med. 42, 3413–3442. doi: 10.1002/sim.9812

Wang L., Jiang Y., Qi H. (2020). Marine dissolved oxygen prediction with tree tuned deep neural network. IEEE Access 8, 182431–182440. doi: 10.1109/ACCESS.2020.3028863

Wang Y., Yuan X., Bi H., Ren Y., Liang Y., Li C., et al. (2023a). Understanding arctic sea ice thickness predictability by a markov model. J. Climate 36, 1–40. doi: 10.1175/JCLI-D-22–0525.1

Wang Y., Yuan X., Ren Y., Bushuk M., Shu Q., Li C., et al. (2023b). Subseasonal prediction of regional antarctic sea ice by a deep learning model. Geophys. Res. Lett. 50, e2023GL104347. doi: 10.1029/2023GL104347

Xu W., Fleddum A. L., Shin P. K., Sun J. (2022). Impact of and recovery from seabed trawling in soft-bottom benthic communities under natural disturbance of summer hypoxia: A case study in subtropical hong kong. Front. Mar. Sci. 9. doi: 10.3389/fmars.2022.1010909

Yang W., Li Z., Liu M., Lu Y., Cao K., Maciejewski R., et al. (2020). “Diagnosing concept drift with visual analytics,” in 2020 IEEE conference on visual analytics science and technology (VAST) (IEEE). IEEE Computer Soc: Los Alamitos, CA, USA. 12–23. doi: 10.1109/VAST50239.2020.00007

Ye M., Li B., Nie J., Wen Q., Wei Z., Yang L.-L. (2023). Graph convolutional network assisted sst and chl-a prediction with multi-characteristics modeling of spatio-temporal evolution. IEEE Trans. Geosci. Remote Sens . 61. doi: 10.1109/TGRS.2023.3330517

Ye M., Nie J., Liu A., Wang Z., Huang L., Tian H., et al. (2021). Multi-year enso forecasts using parallel convolutional neural networks with heterogeneous architecture. Front. Mar. Sci. 8, 717184. doi: 10.3389/fmars.2021.717184

Zhang G., Hóu Z., Huang Y., Shi J., Bride H., Dong J. S., et al. (2023). Extracting optimal explanations for ensemble trees via automated reasoning. Appl. Intell. 53, 14371–14382. doi: 10.1007/s10489-022-04180-1

Keywords: interpretability, high-precision, dissolved oxygen, causal discovery, drift detection, RuleFit, SHAP, LIME

Citation: Li X, Liu Z, Yang Z, Meng F and Song T (2024) A high-precision interpretable framework for marine dissolved oxygen concentration inversion. Front. Mar. Sci. 11:1396277. doi: 10.3389/fmars.2024.1396277

Received: 05 March 2024; Accepted: 10 May 2024; Published: 31 May 2024.

Reviewed by:

Copyright © 2024 Li, Liu, Yang, Meng and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Tao Song, [email protected]

This article is part of the Research Topic

Deep Learning for Marine Science, Volume II

The Concept and Implications of Colorblind Racism

This essay about colorblind racism explains how the ideology of ignoring race to promote equality can perpetuate racial injustice. It discusses how colorblindness emerged in the post-civil rights era and argues that treating everyone equally without acknowledging race overlooks systemic inequalities. The essay highlights how colorblind racism denies racial differences, avoids discussions about race, and critiques policies like affirmative action, which aim to address inequality. It emphasizes the need to recognize and address systemic racism, validate the experiences of people of color, and foster open conversations about race to achieve genuine equality and justice.

How it works

The concept of colorblind racism delineates a modern iteration of racial prejudice that feigns ignorance toward the significance of race in human existence and choices. This ideology advocates for the notion that eradicating discrimination entails treating all individuals with parity, irrespective of their racial, ethnic, or chromatic attributes. Nonetheless, this perspective often overlooks the entrenched systemic disparities persisting within society, inadvertently perpetuating racial injustice.

The notion of colorblindness in societal frameworks gained prominence notably during the post-civil rights era, notably in the United States.

Proponents argue that by averting acknowledgment of race, society can transcend historical injustices and foster a genuinely equitable environment. However, critics contend that this viewpoint oversimplifies the matter and fails to acknowledge the pervasive entrenchment of race and racism within societal structures. It posits that ceasing to recognize race would result in the eradication of racial issues, a notion starkly distant from reality.

Colorblind racism materializes in diverse manifestations. One prevalent manifestation is the repudiation of racial disparities, coupled with the assertion that everyone enjoys equal opportunities. This stance disregards the historical context of racial bias and the persistent disparities evident in realms such as education, employment, housing, and criminal justice. For instance, the notion that meritocracy singularly propels success overlooks systemic impediments disproportionately affecting people of color. It neglects the enduring ramifications of historical discrimination and the perpetuation of inequitable policies and practices disadvantaging specific racial demographics.

Another manifestation is the evasion of conversations regarding race and racism. While dialogues on race may induce discomfort, evading them does not resolve the underlying issues; rather, it complicates their resolution. When individuals espouse colorblindness, they frequently dismiss the lived experiences of individuals of color, thereby trivializing their struggles and viewpoints. This engenders a deficit of comprehension and empathy, exacerbating the challenge of attaining genuine equity.

Moreover, colorblind racism is discernible in the critique of policies crafted to redress racial inequality, such as affirmative action. Opponents argue that such policies constitute reverse discrimination, advocating for a uniform treatment of all individuals. However, this stance disregards historical instances of unequal opportunities and underscores the necessity of proactive measures to level the playing field. By denouncing these endeavors as unjust, colorblind racism perpetuates the status quo, impeding progress toward equity.

The ideology of colorblindness also influences interpersonal dynamics. For instance, individuals asserting to “not see color” may perceive themselves as non-racist. However, this assertion proves problematic as it disregards the import of racial identity and the repercussions of racism on individuals’ lives. It obstructs substantive discussions on race and impedes the formulation of policies and practices addressing racial disparities.

Combatting colorblind racism necessitates a paradigm shift. It demands acknowledgment of the salience of race and the pervasive nature of systemic racism. This entails heeding and validating the experiences of individuals of color, recognizing the historical and structural factors underpinning inequality, and actively dismantling these barriers. It also involves fostering transparent discussions on race, fostering enhanced comprehension and empathy.

In educational contexts, this could entail integrating diverse viewpoints into curricula and establishing forums for students to engage in dialogues on race and racism. In workplaces, it might entail implementing policies fostering diversity and inclusivity and rectifying disparities in recruitment, advancement, and compensation. On a broader societal scale, it requires endorsing policies addressing systemic disparities and advocating for justice and equity.

In conclusion, colorblind racism epitomizes a multifaceted and pernicious manifestation of racism that, albeit ostensibly progressive, ultimately perpetuates racial disparities. By disregarding the realities of race and racism, it precludes society from confronting the root causes of discrimination, impeding progress toward authentic equity. Overcoming colorblind racism necessitates recognition and confrontation of the systemic underpinnings of racial disparities, along with active endeavors to promote racial justice and equity. Only through acknowledging the import of race and addressing its repercussions can society aspire to forge a more just and equitable future for all.

Cite this page

The Concept and Implications of Colorblind Racism. (2024, Jun 01). Retrieved from https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/

"The Concept and Implications of Colorblind Racism." PapersOwl.com , 1 Jun 2024, https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/

PapersOwl.com. (2024). The Concept and Implications of Colorblind Racism . [Online]. Available at: https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/ [Accessed: 3 Jun. 2024]

"The Concept and Implications of Colorblind Racism." PapersOwl.com, Jun 01, 2024. Accessed June 3, 2024. https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/

"The Concept and Implications of Colorblind Racism," PapersOwl.com , 01-Jun-2024. [Online]. Available: https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/. [Accessed: 3-Jun-2024]

PapersOwl.com. (2024). The Concept and Implications of Colorblind Racism . [Online]. Available at: https://papersowl.com/examples/the-concept-and-implications-of-colorblind-racism/ [Accessed: 3-Jun-2024]

Don't let plagiarism ruin your grade

Hire a writer to get a unique paper crafted to your needs.

Our writers will help you fix any mistakes and get an A+!

Please check your inbox.

You can order an original essay written according to your instructions.

Trusted by over 1 million students worldwide

1. Tell Us Your Requirements

2. Pick your perfect writer

3. Get Your Paper and Pay

Hi! I'm Amy, your personal assistant!

Don't know where to start? Give me your paper requirements and I connect you to an academic expert.

short deadlines

100% Plagiarism-Free

Certified writers

• Share full article

Advertisement

Subscriber-only Newsletter

The Ethicist

Can i use a.i. to grade my students’ papers.

The magazine’s Ethicist columnist on artificial intelligence platforms, and whether it’s hypocritical for teachers to use these tools while forbidding students from doing the same.

By Kwame Anthony Appiah

I am a junior-high-school English teacher. In the past school year, there has been a significant increase in students’ cheating on writing assignments by using artificial intelligence. Our department feels that 13-year-old students will only become better writers if they practice and learn from the successes and challenges that come with that.

Recently our department tasked students with writing an argumentative essay, an assignment we supported by breaking down the process into multiple steps. The exercise took several days of class time and homework to complete. All of our students signed a contract agreeing not to use A.I. assistance, and parents promised to support the agreement by monitoring their children when they worked at home. Yet many students still used A.I.

Some of our staff members uploaded their grading rubric into an A.I.-assisted platform, and students uploaded their essays for assessment. The program admittedly has some strengths. Most notable, it gives students writing feedback and the opportunity to edit their work before final submission. The papers are graded within minutes, and the teachers are able to transfer the A.I. grade into their roll book.

I find this to be hypocritical. I spend many hours grading my students’ essays. It’s tedious work, but I feel that it’s my responsibility — if a student makes an effort to complete the task, they should have my undivided attention during the assessment process.

Here’s where I struggle: Should I embrace new technology and use A.I.-assisted grading to save time and my sanity even though I forbid my students from using it? Is it unethical for teachers to ask students not to use A.I. to assist their writing but then allow an A.I. platform to grade their work? — Name Withheld

From the Ethicist:

You have a sound rationale for discouraging your students from using A.I. to draft their essays. As with many other skills, writing well and thinking clearly will improve through practice. By contrast, you already know how to grade papers; you don’t need the practice.

What matters is whether an A.I.-assisted platform can reliably appraise and diagnose your students’ writing, providing the explanation and guidance these students need to improve. In theory, such tools — and I see that there are several on the market, including from major educational publishers — have certain advantages. The hope is that they can grade without inconsistency, without getting tired, without being affected by the expectations that surely affect those of us who hand-grade student work.

I notice you haven’t raised concerns about whether the platform provides reliable assessments; you’ll have to decide if it does. (If it isn’t quite up to snuff, it might become so in a year or two, so your question will persist.) Provided the platform does a decent job of assessment, though, I don’t see why you must do it all yourself. You should review the A.I.-annotated versions of your students’ writing, check that you agree with the output, and make notes of issues to bring up in class. But time saved in evaluating the papers might be better spent on other things — and by “better,” I mean better for the students. There are pedagogical functions, after all, that only you can perform.

In sum: It’s not hypocritical to use A.I. yourself in a way that serves your students well, even as you insist that they don’t use it in a way that serves them badly.

Readers Respond

The previous question was from a reader who asked about professional boundaries. He wrote: “I am a retired, married male psychiatrist. A divorced female former patient of mine contacted me recently, 45 years after her treatment ended. Would it be OK to correspond with her by email? Or is this a case of ‘once a patient, always a patient?’”

In his response, the Ethicist noted: “The relevant professional associations tend to have strictures that are specifically about sexual relationships with former patients. … In light of the potential for exploitation within the therapist-patient relationship, these rules are meant to maintain clear boundaries, protect patient welfare, uphold the integrity of the profession and eliminate any gray areas that could lead to ethical breaches. But though you do mention her marital status, and yours, you’re just asking about emailing her — about establishing friendly relations. The question for you is whether she might be harmed by this, whether whatever knowledge or trust gained from your professional relationship would shadow a personal one. Yes, almost half a century has elapsed since your professional relationship, but you still have to be confident that a correspondence with her clears this bar. If it does, you may email with a clear conscience.” ( Reread the full question and answer here. )

As always, I agree with the Ethicist. I would add that the letter writer’s former patient doesn’t realize that the therapist is actually two different people — the professional and the regular person underneath. Therapists portray their professional selves to their clients. The former client may be disappointed upon meeting the therapist outside of the professional context. Additionally, the feelings she has toward the therapist may be based on transference, and they would need to address that. — Annemarie

I am a clinical psychologist. While the Ethicist’s description of professional ethical boundaries is correct, there is more to the story, and I disagree with his conclusion. A very big question here is why this former patient contacted him after 45 years. That is a question that is best explored and answered within the context of a therapeutic relationship. He would be well- advised to respond in a kind and thoughtful way to convey the clear message that he is not available for ongoing communication, and he should suggest that she consult with another therapist if she feels that would be helpful. — Margaret

In my case, it was the therapist who reached out to me, seeking to establish a friendship several years after our sessions ended. I was surprised, but he shared that he had since experienced a similar personal tragedy to one I had explored with him in sessions. Since it had been several years since we saw each other professionally, I responded. There was never any hint of romantic or sexual interest. Still, as he continued to reach out to me, clearly desiring a friendship, it never felt right to me. It did feel unprofessional, as his knowledge of me was borne out of a relationship meant to be professional, never personal, as warmly as we might have felt during our sessions. I ended up being disappointed in him for seeking out my friendship. — Liam

I am a (semi)retired psychiatrist who has been practicing since 1974. In my opinion, “once a patient, always a patient” is correct. Establishing any type of personal relationship with a former patient could undo progress the patient may have made in treatment, and is a slippery slope toward blatantly unethical behavior. As psychiatrists, our responsibility is to work with patients in confronting and resolving issues that are preventing them from having a reality-based perception of their life. With such an outlook, they are more capable of establishing satisfying relationships with others. An ethical psychiatrist is not in the business of providing such satisfaction to his or her patients. — Roger

I think there is a difference between being friendly and being friends with a former client. As someone who used to attend therapy with a therapist I think dearly of, she made it clear to me that it was OK to send her emails with life updates after our therapeutic relationship ended. But beyond that, I think it would be inappropriate and uncomfortable to pursue a friendship with her, and vice versa, because of the patient-provider relationship that we previously had and the power dynamic that existed between us. The letter writer didn’t share the content of the email his former patient sent to him, but if it’s just a friendly life update, I think it’s fine to write back and thank her for sharing. Beyond that, I feel like it would be unprofessional to meet or pursue a deeper relationship. — Meghan

Kwame Anthony Appiah is The New York Times Magazine’s Ethicist columnist and teaches philosophy at N.Y.U. His books include “Cosmopolitanism,” “The Honor Code” and “The Lies That Bind: Rethinking Identity.” To submit a query: Send an email to [email protected]. More about Kwame Anthony Appiah

Help | Advanced Search

Computer Science > Cryptography and Security

Title: voice jailbreak attacks against gpt-4o.

Abstract: Recently, the concept of artificial assistants has evolved from science fiction into real-world applications. GPT-4o, the newest multimodal large language model (MLLM) across audio, vision, and text, has further blurred the line between fiction and reality by enabling more natural human-computer interactions. However, the advent of GPT-4o's voice mode may also introduce a new attack surface. In this paper, we present the first systematic measurement of jailbreak attacks against the voice mode of GPT-4o. We show that GPT-4o demonstrates good resistance to forbidden questions and text jailbreak prompts when directly transferring them to voice mode. This resistance is primarily due to GPT-4o's internal safeguards and the difficulty of adapting text jailbreak prompts to voice mode. Inspired by GPT-4o's human-like behaviors, we propose VoiceJailbreak, a novel voice jailbreak attack that humanizes GPT-4o and attempts to persuade it through fictional storytelling (setting, character, and plot). VoiceJailbreak is capable of generating simple, audible, yet effective jailbreak prompts, which significantly increases the average attack success rate (ASR) from 0.033 to 0.778 in six forbidden scenarios. We also conduct extensive experiments to explore the impacts of interaction steps, key elements of fictional writing, and different languages on VoiceJailbreak's effectiveness and further enhance the attack performance with advanced fictional writing techniques. We hope our study can assist the research community in building more secure and well-regulated MLLMs.

Submission history

Access paper:.

• HTML (experimental)
• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .