Plagiarism Policy

Originality Requirement

Submissions must be original, properly cited, and not under active consideration elsewhere. Prior posting as a preprint is acceptable when disclosed, but duplicate journal submission is not acceptable.

Similarity Screening

The editorial office may use similarity-checking tools, manual source comparison, reviewer alerts, and editorial judgement. A high similarity score does not automatically prove misconduct; context, quotation, methods reuse, references, and legitimate overlap must be assessed.

Field-Specific Risks

For applied machine learning, editors should pay particular attention to template ML experiments, copied code, undisclosed generated text, duplicated benchmarks, and exaggerated deployment language. These concerns may involve text, images, code, datasets, protocols, media, designs, or supplementary material.

Unacceptable Practices

The journal may reject or retract work involving plagiarism, unattributed paraphrase, image or data reuse without permission, duplicate publication, salami slicing, fabricated references, manipulated peer review, undisclosed third-party writing, or undisclosed generated content.

Text Recycling

Limited reuse of an author's own methods text may be acceptable when the source is cited and the overlap is necessary for reproducibility. Reuse of results, interpretation, literature review, policy argument, case material, or conclusions without transparency is not acceptable.

Process

When concerns arise, authors will normally be asked for an explanation. Serious or unresolved concerns may lead to rejection, correction, retraction, institutional notification, funder notification, or other action consistent with publication ethics guidance.

Reviewer Alerts

Reviewers should alert the editor if they recognise duplicate text, copied figures, repeated datasets, suspicious citations, paper-mill patterns, or undisclosed conflicts. Reviewers should not investigate by contacting authors directly.