Automatic detection of production-sustaining linguistic units
(2025) Traitement de données langagières dynamiques par les outils et méthodes du TAL (DYN-TAL), workshop at CORIA-TALN 2025 — Location: Université d’Aix-Marseille (30.June.2025)
Files
FeltgenGilquin2025.pdf
Open Access - Adobe PDF
- 561.9 KB
Details
- Authors
- Abstract
- Writing production unfolds in ebbs and flows, segmented by pauses into bursts of writing that are believed to match an underlying cognitive consistency of the writing process. To strengthen their writing fluency, writers may rely on linguistic units that sustain the production flow. This contribution aims to offer an NLP method to automatically detect these production-sustaining units. We apply it to a learner corpus of English L2 writing and show that text-structuring markers (connectors and the like) work in an isolated way. On the other hand, this process can be sustained by complementation strategies, such as the reliance on prepositional phrases or infinitive clauses, which elaborate on the existing content.
- Affiliations
Citations
Feltgen, Q., & Gilquin, G. (2025). Automatic detection of production-sustaining linguistic units. In F. Bechet, A.-G. Chifu, K. Pinel-sauvagnat, B. Favre, E. Maes, D. Nurbakova (ed.), Actes de l’atelier Traitement de données langagières dynamiques par les outils et méthodes, CORIA-TALN-RJCRI-RECITAL 2025 (pp. 17-22). https://hdl.handle.net/2078.5/274929