EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Context-aware large language models for ambiguity detection in requirements

Victor Vilhelm Poulsen, Matthias Guertler, Boris Eisenbart, Laura Tomidei, Nathalie Sick

Proceedings of the Design Society · 2026-08 · 6 · 2493-2502

DOI: 10.1017/pds.2026.10607 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the Design Society, online erschienen am 02.07.2026. Verfasst von Victor Vilhelm Poulsen und 4 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Requirements quality shapes engineering design, yet natural language specifications remain vulnerable to ambiguity. We investigate how LLMs support ambiguity detection using a hybrid dataset combining NASA JWST requirements with systematically injected defects. Using auto-extracted domain knowledge, we compare a domain-agnostic baseline with a context-aware approach. Incorporating domain knowledge helps LLMs better distinguish genuinely ambiguous requirements from acceptable ones, highlighting the potential of context-aware AI assistants for requirements engineering and early-stage design.

Zitieren

APA

Poulsen, V. V., Guertler, M., Eisenbart, B., Tomidei, L., & Sick, N. (2026). Context-aware large language models for ambiguity detection in requirements. Proceedings of the Design Society, 6, 2493-2502. https://doi.org/10.1017/pds.2026.10607

Vancouver

Poulsen VV, Guertler M, Eisenbart B, Tomidei L, Sick N. Context-aware large language models for ambiguity detection in requirements. Proceedings of the Design Society. 2026;6:2493-2502. doi:10.1017/pds.2026.10607

BibTeX
@article{Poulsen2026,
  author = {Poulsen, Victor Vilhelm and Guertler, Matthias and Eisenbart, Boris and Tomidei, Laura and Sick, Nathalie},
  title = {Context-aware large language models for ambiguity detection in requirements},
  journal = {Proceedings of the Design Society},
  year = {2026},
  volume = {6},
  pages = {2493--2502},
  doi = {10.1017/pds.2026.10607}
}
Kostenlos starten →