Attributing Fair Decisions with Attention Interventions

TL;DR

This paper introduces an attention-based model that provides fair decision attributions and a post-processing bias mitigation strategy, enhancing transparency and fairness in AI systems across tabular and textual data.

Contribution

The paper presents a novel attention intervention framework for attribution and fairness, along with a post-processing bias mitigation method, applicable to diverse data types.

Findings

Effective identification of features responsible for fairness and performance.

Successful bias mitigation demonstrated on tabular and textual datasets.

Attention interventions improve interpretability and fairness of AI models.

Abstract

The widespread use of Artificial Intelligence (AI) in consequential domains, such as healthcare and parole decision-making systems, has drawn intense scrutiny on the fairness of these methods. However, ensuring fairness is often insufficient as the rationale for a contentious decision needs to be audited, understood, and defended. We propose that the attention mechanism can be used to ensure fair outcomes while simultaneously providing feature attributions to account for how a decision was made. Toward this goal, we design an attention-based model that can be leveraged as an attribution framework. It can identify features responsible for both performance and fairness of the model through attention interventions and attention weight manipulation. Using this attribution framework, we then design a post-processing bias mitigation strategy and compare it with a suite of baselines. We demonstrate the versatility of our approach by conducting experiments on two distinct data types, tabular and textual.