Visual Analysis of Discrimination in Machine Learning

TL;DR

This paper introduces DiscriLens, a visual analytics tool that helps identify and interpret discrimination in machine learning models through interactive visualizations and causal analysis, aiding fairness assessment.

Contribution

It presents a novel visualization approach combining Euler diagrams and matrices to analyze discriminatory patterns in machine learning, supported by causal modeling and rule mining.

Findings

Users can interpret discrimination visually quickly and accurately.

DiscriLens effectively guides understanding and reduction of algorithmic bias.

The tool facilitates comprehensive discrimination analysis in critical applications.

Abstract

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.