Plagiarism Screening and AI Policy

The Journal of Vocational in Aquaculture (JAVA) is committed to publishing original, honest, and verifiable research. This policy explains how JAVA screens manuscripts for plagiarism, who qualifies as an author, and how artificial intelligence (AI) tools may and may not be used. It follows the guidelines of the Committee on Publication Ethics (COPE).

Plagiarism screening with Turnitin
Forms of plagiarism
How suspected plagiarism is handled
Authorship
Use of AI by authors
AI-generated images and data
Use of AI by reviewers and editors
Non-compliance

1. Plagiarism screening with Turnitin

All manuscripts submitted to JAVA are screened for similarity with Turnitin:

before peer review, during the editor's initial check; and
again after the final revision, before the manuscript is accepted for publication.

Screening tool	Turnitin
When	Before peer review and before acceptance
Overall similarity accepted	Up to 20%
Similarity with any single source	Up to 3%
Excluded from the score	Reference list, properly quoted and cited text, and matches of fewer than 10 words

The similarity score is a guide, not an automatic decision. The editor reads every Turnitin report to judge whether the overlap is acceptable, for example standard wording in a Materials and Methods section, or whether it is plagiarism. Similarity with the authors' own thesis, dissertation, or preprint is acceptable if it is disclosed in the cover letter.

2. Forms of plagiarism

JAVA considers the following to be plagiarism or redundant publication:

copying text, data, figures, or ideas from another source without proper citation;
paraphrasing another work closely, including with paraphrasing or AI tools, without citation;
reusing substantial parts of the authors' own published work without citation (self-plagiarism or text recycling);
submitting the same study, or substantially the same results, to more than one journal (duplicate or redundant publication); and
translating another published work and presenting it as new.

3. How suspected plagiarism is handled

Similarity within the limits	The manuscript continues to peer review or acceptance.
Similarity above the limits, without evidence of misconduct	The manuscript is returned to the authors to rewrite and cite properly, and is screened again on resubmission.
Clear plagiarism or duplicate submission	The manuscript is rejected, and the authors' institution may be informed.
Plagiarism found after publication	The case is investigated following the COPE flowcharts. Depending on the outcome, JAVA publishes a correction or retracts the article.

4. Authorship

Everyone listed as an author must:

have made a substantial contribution to the study design, data collection, analysis, or interpretation;
have drafted or critically revised the manuscript;
have approved the final version; and
agree to be accountable for the work.

People who contributed in other ways, for example by providing facilities, technical help, or funding, should be named in the Acknowledgements. Gift, guest, and ghost authorship are not accepted. Changes to the author list after submission require the written agreement of all authors. Authors are encouraged to describe each author's contribution using the CRediT taxonomy.

5. Use of AI by authors

AI tools, including ChatGPT, Gemini, Copilot, and other large language models, cannot be listed as authors, because they cannot take responsibility for the work. Authors remain fully responsible for the accuracy, originality, and integrity of everything in their manuscript.

Permitted with disclosure

Improving language, grammar, and readability (for example Grammarly, DeepL, or QuillBot).
Summarising or reorganising the authors' own text, provided the authors check the result.
AI-based tools that are part of the research method itself, described in the Materials and Methods.

Not permitted

Generating research ideas, hypotheses, results, interpretations, or conclusions.
Creating, altering, or fabricating data.
Generating references or citations. Every reference must be checked against the original source.
Paraphrasing other authors' work to hide its origin.

AI Declaration

Every manuscript must include an AI Declaration section, placed before the References, using one of the following statements:

If AI tools were used:
"During the preparation of this manuscript, the author(s) used [NAME OF TOOL] for [PURPOSE, e.g., language editing]. The author(s) reviewed and edited all AI-assisted content and take full responsibility for the content of the publication."

If no AI tools were used:
"The author(s) did not use any artificial intelligence tools in the preparation of this manuscript."

6. AI-generated images and data

Figures, photographs, and graphs that show research results must not be created or altered with generative AI.
Adjusting brightness, contrast, or colour balance is acceptable if it is applied to the whole image and does not hide or change any information.
If AI is part of the research method (for example image analysis to count or measure fish), describe the tool, its version, and its settings in the Materials and Methods.
Editors may ask for the original, unprocessed images or raw data at any stage.

7. Use of AI by reviewers and editors

Manuscripts under review are confidential. Reviewers and editors must not upload a manuscript, or any part of it, to generative AI tools, because this may breach the authors' confidentiality and data rights. Reviewers must write their own review reports. They may use AI tools only to improve the language of a report they wrote themselves, and must not use them to assess the manuscript.

8. Non-compliance

The editors may ask authors to clarify how AI tools were used, or to provide Turnitin-related explanations, at any stage. Manuscripts that breach this policy may be rejected. Published articles may be corrected or retracted following COPE guidance.

Questions about this policy can be sent to java.ikn@apps.ipb.ac.id.