Guillaume Cabanac
Institut de Recherche en Informatique de Toulouse (IRIT)
Fake Science: Misconduct Galore and Proposed Counterattacks
Imagine reading invulnerable framework, glucose bigotry, and nucleic corrosive* in a scientific paper published in an established journal with Impact Factor from an haute couture publisher such as Elsevier, Wiley, Springer... Well, we identified more than 14k such problematic papers featuring ‘tortured phrases’ produced by paraphrasing tools to deceive plagiarism detection. In 2021, this issue affected 3 every 10k papers. Scammers produce such unreliable papers in bulk to appear hyper productive; they organise citation rings to increase their h-index.
This talk presents how sleuths strive to decontaminate the literature from this digital pollution. A handful of professional scientists and hobbyists re-assess the suspect papers flagged by the ‘Problematic Paper Screener,’ contribute pro bono post-publication peer reviews on PubPeer, and eventually manage to get publishers retract 3000+ papers, mostly in engineering and biomedicine. The talk will also cover the AI misuse to generate text and images like a chimeric rat or lungs.
*instead of immune system, glucose intolerance, and nucleic acid.
Selected publications
- Cabanac, G. (2021). This digital-hygiene routine will protect your scholarship [World View: A personal take on science and society]. In Nature (Vol. 598, Issue 7882, p. 541). https://doi.org/10.1038/d41586-021-02901-z
- Cabanac, G., & Labbé, C. (2021). Prevalence of nonsensical algorithmically generated papers in the scientific literature. In Journal of the Association for Information Science and Technology (Vol. 72, Issue 12, pp. 1461–1476). https://doi.org/10.1002/asi.24495
- Cabanac, G., Labbé, C., & Magazinov, A. (2021). Tortured phrases: A dubious writing style emerging in science. Evidence of critical issues affecting established journals. arXiv preprint. https://arxiv.org/abs/2107.06751 (supplementary materials)
- Cabanac, G., Labbé, C., & Magazinov, A. (2022). “Bosom peril” is not “breast cancer”: How weird computer-generated phrases help researchers find scientific publishing fraud. Bulletin of the Atomic Scientists
- Cabanac, G. (2022). Decontamination of the scientific literature. arXiv preprint. https://doi.org/10.48550/arXiv.2210.15912