{"paper":{"title":"User Reviews as a Source for Usability Requirements: A Precursor Study on Using Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Large language models can identify usability requirements in user reviews with F-scores comparable to human raters when the prompt is well designed.","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Cedric Wellhausen, Kurt Schneider, Laura Reinhardt","submitted_at":"2026-05-12T19:05:04Z","abstract_excerpt":"It is known that user-centered approaches to requirements engineering in general lead to a better suited product for the end-users. LLM4RE provides promising approaches to support the requirements elicitation process (e.g. classification of requirements). Previous approaches focus on Machine-Learning (ML) or Deep-Learning (DL) aspects, which require intensive training with a large amount of manually labeled data. LLMs, on the other hand, are pre-trained on large amounts of user-generated text data, enabling a user-centric workflow to analyze requirements. In this paper, we explore the possibil"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We determine that LLMs are generally able to recognize usability as a non-functional requirement in user reviews, in terms of their F-score, but the performance and reliability is strongly dependent on the prompt.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the human raters provide consistent ground truth labels and that the prompt developed on this dataset will produce reliable results on new reviews or different LLMs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"LLMs can detect usability content in user reviews with F-scores comparable to humans, though performance depends strongly on prompt design.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Large language models can identify usability requirements in user reviews with F-scores comparable to human raters when the prompt is well designed.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"183e3b8cdaf39803031a19a27a7c6a14ca630a9b1eff187e2445504dd621fc21"},"source":{"id":"2605.12657","kind":"arxiv","version":1},"verdict":{"id":"3f555087-0e57-46da-a353-9e61c970e696","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T20:20:58.447696Z","strongest_claim":"We determine that LLMs are generally able to recognize usability as a non-functional requirement in user reviews, in terms of their F-score, but the performance and reliability is strongly dependent on the prompt.","one_line_summary":"LLMs can detect usability content in user reviews with F-scores comparable to humans, though performance depends strongly on prompt design.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the human raters provide consistent ground truth labels and that the prompt developed on this dataset will produce reliable results on new reviews or different LLMs.","pith_extraction_headline":"Large language models can identify usability requirements in user reviews with F-scores comparable to human raters when the prompt is well designed."},"references":{"count":26,"sample":[{"doi":"10.1109/rew.2017.76","year":2017,"title":"E. Bakiu, E. Guzman, Which feature is unusable? detecting usability and user experience issues from user reviews, in: 2017 IEEE 25th International Requirements Engineering Conference Workshops (REW), ","work_id":"431bed5b-4d1b-430a-b24e-e07ef04bf16c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.33540/3091","year":2025,"title":"Groen, Crowd-Based Requirements Engineering, Doctoral thesis 2 (research not uu / graduation uu), Universiteit Utrecht, 2025","work_id":"733c5b63-f823-4e46-9598-154e238bfaa1","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1145/3444689","year":2021,"title":"L. Zhao, W. Alhoshan, A. Ferrari, K. J. Letsholo, M. A. Ajagbe, E.-V. Chioasca, R. T. Batista-Navarro, Natural language processing for requirements engineering: A systematic mapping study, ACM Comput.","work_id":"a106466d-85ec-4c4f-b8c6-618579be471b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Large language models (llms) for requirements engineering (re): A systematic literature review","work_id":"2eb74df0-3d00-48e2-8142-5f447e9d5188","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/rew57809.2023.00024","year":2023,"title":"Revisiting the Performance-Explainability Trade-Off in Explainable Artificial Intelligence (XAI)","work_id":"7e50a9a4-df1e-4f4a-afd3-fd45d6f41c29","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":26,"snapshot_sha256":"7f68631d2036a3ef6f41d005349b64236ff5b3a9cc9fc1625b47220b1cac640e","internal_anchors":3},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}