Disability-First AI Dataset Annotation: Co-designing Stuttered Speech Annotation Guidelines with People Who Stutter Conference: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems; April 13–17, 2026; Barcelona, Spain Proceedings of the 2026 CHI Conference on Human Factors in
Xinru Tang, Jingjin Li, Shaomei Wu · 2026
Read Disability-First AI Dataset Annotation: Co-designing Stuttered Speech Annotation Guidelines with People Who Stutter Conference: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems; April 13–17, 2026; Barcelona, Spain Proceedings of the 2026 CHI Conference on Human Factors in by Xinru Tang, Jingjin Li, Shaomei Wu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-10403.txt.
Description
Data sets meant to include disabled people often get their labels from crowdworkers without disability-specific know-how, yielding inconsistent or inaccurate labels. A 2026 paper led by Xinru Tang, with two co-authors, examines this problem in speech data from people who stutter. Stuttering varies widely, and views differ on how it shows, so labeling and writing down stuttered speech stays hard even for trained experts. Through interviews and co-design workshops with people who stutter and domain experts, the authors name labeling challenges and build practices weaving stutterers' lived experience into labeling.
The findings highlight the value of lived knowledge for data-set quality, while revealing a tension: disability experiences are complex, but fixed labels are rigid. The paper closes by urging disability-first, multiplicity-aware ways of reading data across the AI pipeline, with labeling designed with disabled people rather than merely about them.
Catalog record
| CREATOR | Xinru Tang, Jingjin Li, Shaomei Wu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 102 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-10403