DBAILGMar 23, 2018

Datasheets for Datasets

arXiv:1803.09010v82978 citations
Originality Highly original
AI Analysis

This addresses a foundational problem for the machine learning community by promoting better communication and reducing risks in high-stakes domains, though it is an incremental step in standardization.

The paper tackles the lack of standardized dataset documentation in machine learning by proposing datasheets for datasets, analogous to electronics datasheets, to document motivation, composition, and uses, aiming to improve transparency and accountability.

The machine learning community currently has no standardized process for documenting datasets, which can lead to severe consequences in high-stakes domains. To address this gap, we propose datasheets for datasets. In the electronics industry, every component, no matter how simple or complex, is accompanied with a datasheet that describes its operating characteristics, test results, recommended uses, and other information. By analogy, we propose that every dataset be accompanied with a datasheet that documents its motivation, composition, collection process, recommended uses, and so on. Datasheets for datasets will facilitate better communication between dataset creators and dataset consumers, and encourage the machine learning community to prioritize transparency and accountability.

Code Implementations25 repos

Data from Papers with Code (CC-BY-SA-4.0)

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes