Skip to yearly menu bar Skip to main content


Oral

Position: Measure Dataset Diversity, Don't Just Claim It

Dora Zhao · Jerone Andrews · Orestis Papakyriakopoulos · Alice Xiang

Straus 1-3
award Best Paper
[ ] [ Visit Oral 3E Data and Society ]
Wed 24 Jul 1:30 a.m. — 1:45 a.m. PDT

Abstract:

Machine learning (ML) datasets, often perceived as neutral, inherently encapsulate abstract and disputed social constructs. Dataset curators frequently employ value-laden terms such as diversity, bias, and quality to characterize datasets. Despite their prevalence, these terms lack clear definitions and validation. Our research explores the implications of this issue by analyzing "diversity" across 135 image and text datasets. Drawing from social sciences, we apply principles from measurement theory to identify considerations and offer recommendations for conceptualizing, operationalizing, and evaluating diversity in datasets. Our findings have broader implications for ML research, advocating for a more nuanced and precise approach to handling value-laden properties in dataset construction.

Chat is not available.