Studying Up Machine Learning Data: Why Talk About Bias When We Mean Power?

TL;DR

This paper advocates shifting from bias-focused to power-aware perspectives in ML dataset research, emphasizing historical, social, and labor contexts to better understand data's societal impact.

Contribution

It introduces a power-aware framework for studying ML datasets, expanding beyond bias to include social, historical, and labor considerations in data analysis.

Findings

Highlighting limitations of bias-only approaches

Emphasizing importance of social context in data quality

Proposing expanded transparency in dataset documentation

Abstract

Research in machine learning (ML) has primarily argued that models trained on incomplete or biased datasets can lead to discriminatory outputs. In this commentary, we propose moving the research focus beyond bias-oriented framings by adopting a power-aware perspective to "study up" ML datasets. This means accounting for historical inequities, labor conditions, and epistemological standpoints inscribed in data. We draw on HCI and CSCW work to support our argument, critically analyze previous research, and point at two co-existing lines of work within our community -- one bias-oriented, the other power-aware. This way, we highlight the need for dialogue and cooperation in three areas: data quality, data work, and data documentation. In the first area, we argue that reducing societal problems to "bias" misses the context-based nature of data. In the second one, we highlight the corporate forces and market imperatives involved in the labor of data workers that subsequently shape ML datasets. Finally, we propose expanding current transparency-oriented efforts in dataset documentation to reflect the social contexts of data design and production.