Refining Dataset Documentation: A Two-Year Journey to Improve AI Data Transparency

Authors: (1) TIMNIT GEBRU, Black in AI; (2) JAMIE MORGENSTERN, University of Washington; (3) BRIANA VECCHIONE, Cornell University; (4) JENNIFER WORTMAN VAUGHAN, Microsoft Research; (5) HANNA WALLACH, Microsoft Research; (6) HAL DAUMÉ III, Microsoft Research; University of Maryland; (7) KATE CRAWFORD, Microsoft Research. Table of Links Introduction Development Process Questions and Workflow Impact and Challenges Acknowledgments and References Appendix 2 Development Process We refined the questions and workflow provided in the next section over a period of roughly two years, incorporating many rounds of feedback. First, leveraging our own experiences as researchers with diverse backgrounds working in different domains and institutions, we drew on our knowledge of dataset characteristics, unintentional misuse, unwanted societal biases, and other issues to produce an initial set of questions designed to elicit information about these topics. We then “tested” these questions by creating example datasheets for two widely used datasets: Labeled Faces in the Wild [16] and Pang and Lee’s polarity dataset [22]. We chose these datasets in large part because their creators provided exemplary documentation, allowing us to easily find the answers to many of the questions. While creating these example datasheets, we found gaps in the questions, as well as redundancies and lack of clarity. We therefore refined the questions and distributed them to product teams in two major US-based technology companies, in some cases helping teams to create datasheets for their datasets and observing where the questions did not achieve their intended objectives. Contemporaneously, we circulated an initial draft of this paper to colleagues through social media and on arXiv (draft posted 23 March 2018). Via these channels we received extensive comments from dozens of researchers, practitioners, and policy makers. We also worked with a team of lawyers to review the questions from a legal perspective. We incorporated this feedback to yield the questions and workflow provided in the next section: We added and removed questions, refined the content of the questions, and reordered the questions to better match the key stages of the dataset lifecycle. Based on our experiences with product teams, we reworded the questions to discourage yes/no answers, added a section on “Uses,” and deleted a section on “Legal and Ethical Considerations.” We found that product teams were more likely to answer questions about legal and ethical considerations if they were integrated into sections about the relevant stages of the dataset lifecycle rather than grouped together. Finally, following feedback from the team of lawyers, we removed questions that explicitly asked about compliance with regulations, and introduced factual questions intended to elicit relevant information about compliance without requiring dataset creators to make legal judgments. This paper is available on arxiv under CC 4.0 license. Authors: (1) TIMNIT GEBRU, Black in AI; (2) JAMIE MORGENSTERN, University of Washington; (3) BRIANA VECCHIONE, Cornell University; (4) JENNIFER WORTMAN VAUGHAN, Microsoft Research; (5) HANNA WALLACH, Microsoft Research; (6) HAL DAUMÉ III, Microsoft Research; University of Maryland; (7) KATE CRAWFORD, Microsoft Research. Authors: Authors: (1) TIMNIT GEBRU, Black in AI; (2) JAMIE MORGENSTERN, University of Washington; (3) BRIANA VECCHIONE, Cornell University; (4) JENNIFER WORTMAN VAUGHAN, Microsoft Research; (5) HANNA WALLACH, Microsoft Research; (6) HAL DAUMÉ III, Microsoft Research; University of Maryland; (7) KATE CRAWFORD, Microsoft Research. Table of Links Introduction Introduction Development Process Development Process Questions and Workflow Questions and Workflow Impact and Challenges Impact and Challenges Acknowledgments and References Acknowledgments and References Appendix Appendix 2 Development Process We refined the questions and workflow provided in the next section over a period of roughly two years, incorporating many rounds of feedback. First, leveraging our own experiences as researchers with diverse backgrounds working in different domains and institutions, we drew on our knowledge of dataset characteristics, unintentional misuse, unwanted societal biases, and other issues to produce an initial set of questions designed to elicit information about these topics. We then “tested” these questions by creating example datasheets for two widely used datasets: Labeled Faces in the Wild [16] and Pang and Lee’s polarity dataset [22]. We chose these datasets in large part because their creators provided exemplary documentation, allowing us to easily find the answers to many of the questions. While creating these example datasheets, we found gaps in the questions, as well as redundancies and lack of clarity. We therefore refined the questions and distributed them to product teams in two major US-based technology companies, in some cases helping teams to create datasheets for their datasets and observing where the questions did not achieve their intended objectives. Contemporaneously, we circulated an initial draft of this paper to colleagues through social media and on arXiv (draft posted 23 March 2018). Via these channels we received extensive comments from dozens of researchers, practitioners, and policy makers. We also worked with a team of lawyers to review the questions from a legal perspective. We incorporated this feedback to yield the questions and workflow provided in the next section: We added and removed questions, refined the content of the questions, and reordered the questions to better match the key stages of the dataset lifecycle. Based on our experiences with product teams, we reworded the questions to discourage yes/no answers, added a section on “Uses,” and deleted a section on “Legal and Ethical Considerations.” We found that product teams were more likely to answer questions about legal and ethical considerations if they were integrated into sections about the relevant stages of the dataset lifecycle rather than grouped together. Finally, following feedback from the team of lawyers, we removed questions that explicitly asked about compliance with regulations, and introduced factual questions intended to elicit relevant information about compliance without requiring dataset creators to make legal judgments. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv