Datasheets for Datasets

Research on dataset documentation, evaluation and harms in machine learning. Begin with how data was collected and which uses its documentation supports.

View the work and reading plans

How to study this source

• What should a dataset user know before training a model?

  coreOpening overview and the first section that addresses the pathway question. Stop after 30–45 minutes; this is selected reading.

  Research on dataset documentation, evaluation and harms in machine learning. Begin with how data was collected and which uses its documentation supports.

• What should a dataset user know before training a model?

  coreAbstract and Introduction, then the first developed argument or experiment.

  Research on dataset documentation, evaluation and harms in machine learning. Begin with how data was collected and which uses its documentation supports.

• What does it mean to evaluate an AI system?

  coreAbstract, Introduction and the overview of documentation questions

  Evaluation compares a defined task, dataset and measure; those choices affect what success means.

• How can a dataset shape a system’s behaviour?

  coreDatasheets for Datasets: motivation, composition and collection-process questions

  Collection choices affect who and what a model can learn to represent.

Ideas and questions

Read background definitions when a term blocks the argument. Then return to the source and reconstruct its claim in your own words.

Read alongside, read against