What is plagiarism?

Plagiarism is presenting another person’s words, ideas, structure, data or creative work as if it were your own. The decisive issue is attribution—not whether software found a match.

Four forms that are easy to confuse

Direct copying

Words or sentences are reproduced without clear quotation and citation. Changing punctuation or a few words does not make copied structure independent.

Close paraphrase

The vocabulary changes, but the sequence, argument or distinctive expression remains too close to the source. A citation may be necessary, and the wording should reflect your own understanding.

Idea, data or structure plagiarism

Sources are required for more than sentences. An original model, dataset, diagram, argument or unusual chapter structure can also require attribution.

Self-plagiarism

Reusing your own previously submitted or published work without disclosure may breach institutional rules. Whether reuse is permitted depends on the context and policy.

A match is not the same as plagiarism

A checker detects textual similarity. It cannot know whether the quotation is permitted, whether the source is already cited correctly or which academic rules apply. Bibliographies, standard terminology and statutory wording can create legitimate matches.

No universal “safe percentage” exists

A low overall score can conceal one serious uncited passage. A higher score can result from correctly quoted primary material, a long reference list or standard methods wording. Review the nature of each match rather than targeting a number.

What to do before submission

• Mark direct quotations while taking notes.
• Store page numbers and URLs with each source excerpt.
• Write paraphrases without looking sentence-by-sentence at the original.
• Compare each in-text citation with the bibliography.
• Check your institution’s rules for AI tools, collaboration and reuse.