Papers

My research focuses on how to represent academic writing in such a way that it can be used to detect academic misconduct, recommend reviewers, recommend academic collaborations, and more.

Cite Lens: An AI Tool for Detecting Out-of-Scope and Out-of-Context Citations

Jean-Baptiste de la Broise, Frank Sauerburger, Enric Sayas, Dan-Marin Tecu, Sanita Meijere, Milos Cuculovic · Book chapter · 2026

Introduces Cite Lens, an embedding-similarity tool for detecting problematic citation behavior—either misalignment between an article and its references, or between a reference and its citing paragraph. We analyze citation patterns across publishers and topics and show how this approach supports editorial screening against unethical or manipulative citations.

How to pretrain an efficient cross-disciplinary language model: The ScilitBERT use case

Jean-Baptiste de la Broise, Nolwenn Bernard, Jean-Philippe Dubuc, Andrea Perlato, Bastien Latard · 6th International Conference on Information Technology Research (ICITR) · 2021

Presents ScilitBERT, a BERT model pretrained on a cross-disciplinary academic corpus. Half as deep as RoBERTa with much lower pretraining cost, it reaches at least 96% of RoBERTa's accuracy on two downstream academic tasks—showing that dedicated domain pretraining is worthwhile when enough raw text is available.