Visual Analysis of Discrimination in Machine Learning
This addresses fairness issues in critical applications like crime prediction and college admission, but it is incremental as it builds on existing visual analytics and discrimination analysis methods.
The paper tackles the problem of analyzing fairness in machine learning by developing DiscriLens, an interactive visualization tool that identifies discriminatory itemsets using causal modeling and classification rules mining, with a user study showing users can interpret the information quickly and accurately.
The growing use of automated decision-making in critical applications, such as crime prediction and college admission, has raised questions about fairness in machine learning. How can we decide whether different treatments are reasonable or discriminatory? In this paper, we investigate discrimination in machine learning from a visual analytics perspective and propose an interactive visualization tool, DiscriLens, to support a more comprehensive analysis. To reveal detailed information on algorithmic discrimination, DiscriLens identifies a collection of potentially discriminatory itemsets based on causal modeling and classification rules mining. By combining an extended Euler diagram with a matrix-based visualization, we develop a novel set visualization to facilitate the exploration and interpretation of discriminatory itemsets. A user study shows that users can interpret the visually encoded information in DiscriLens quickly and accurately. Use cases demonstrate that DiscriLens provides informative guidance in understanding and reducing algorithmic discrimination.