Sandrine Chausson
@sandachausson.bsky.social
PhD student in Natural Language Processing and Computational Social Science at the University of Edinburgh 🏴
Truly honoured that our paper got such a mention from @lauraknelson.bsky.social! If you’re in need of a versatile and lightweight text classification methodology for sociological research, please consider giving the “Insight-Inference Loop” a read 😊
This article is really good. Among everything coming out around LMs in sociology, this is top of the list for me. They frame text classification at a Natural Language Inference problem, and use BERT(MNLI) to produce an "entailment" score for document-claim pairs. Reproducible, careful, usable work.
The Insight-Inference Loop: Efficient Text Classification via Natural Language Inference and Threshold-Tuning - Sandrine Chausson, Marion Fourcade, David J. Harding, Björn Ross, Grégory Renard, 2025
Modern computational text classification methods have brought social scientists tantalizingly close to the goal of unlocking vast insights buried in text data—f...
journals.sagepub.com
June 24, 2025 at 8:38 PM
Truly honoured that our paper got such a mention from @lauraknelson.bsky.social! If you’re in need of a versatile and lightweight text classification methodology for sociological research, please consider giving the “Insight-Inference Loop” a read 😊