Aspirations and Practice of ML Model Documentation: Moving the Needle with Nudging and Traceability

要旨

The documentation practice for machine-learned (ML) models often falls short of established practices for traditional software, which impedes model accountability and inadvertently abets inappropriate or misuse of models. Recently, model cards, a proposal for model documentation, have attracted notable attention, but their impact on the actual practice is unclear. In this work, we systematically study the model documentation in the field and investigate how to encourage more responsible and accountable documentation practice. Our analysis of publicly available model cards reveals a substantial gap between the proposal and the practice. We then design a tool named DocML aiming to (1) nudge the data scientists to comply with the model cards proposal during the model development, especially the sections related to ethics, and (2) assess and manage the documentation quality. A lab study reveals the benefit of our tool towards long-term documentation quality and accountability.

著者
Avinash Bhat
McGill University, Montreal, Quebec, Canada
Austin Coursey
Vanderbilt University, Nashville, Tennessee, United States
Grace Hu
McGill University, Montreal, Quebec, Canada
Sixian Li
McGill University, Montreal, Quebec, Canada
Nadia Nahar
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Shurui Zhou
University of Toronto, Toronto, Ontario, Canada
Christian Kästner
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Jin L.C. Guo
McGill University, Montreal, Quebec, Canada
論文URL

https://doi.org/10.1145/3544548.3581518

動画

会議: CHI 2023

The ACM CHI Conference on Human Factors in Computing Systems (https://chi2023.acm.org/)

セッション: Transportation and AI/ML

Hall G1
5 件の発表
2023-04-26 18:00:00
2023-04-26 19:30:00