Analysis of mathematical expressions to detect concealed scientific plagiarism

Service providers in the field of plagiarism detection are currently concentrating primarily on the identification of barely concealed forms of plagiarism. In contrast, the more complex research and development effort for tools and services that enable the detection of disguised forms of plagiarism in scientific publications is still being undertaken to a lesser extent. This has recently become much more important as a result of public discussions on artificial text generation using systems such as ChatGPT.

To achieve this goal, we are investigating how potentially suspicious similarities between documents can be detected by analysing mathematical expressions as text- and language-independent features. The project aims to develop methods for detecting disguised scientific plagiarism, such as paraphrases, translations or plagiarized ideas, as they occur especially in the so-called STEM disciplines.

• Duration: April 01, 2022 – September 30, 2025
• Project sponsor: German Research Foundation (DFG)
• Cooperation partner: University of Göttingen
• Responsible for the project: Dr. Moritz Schubotz