Improving Minimax Group Fairness in Sequential Recommendation

TL;DR

This paper applies Distributionally Robust Optimization methods, especially CVaR, to sequential recommenders to improve fairness across diverse user groups, outperforming standard training and handling overlapping groups effectively.

Contribution

It introduces CVaR-based DRO for recommenders, demonstrating improved fairness and overall performance, especially with intersecting user groups.

Findings

CVaR DRO outperforms standard training and other DRO methods.

Group and Streaming DRO are sensitive to group choice.

CVaR effectively handles overlapping user groups.

Abstract

Training sequential recommenders such as SASRec with uniform sample weights achieves good overall performance but can fall short on specific user groups. One such example is popularity bias, where mainstream users receive better recommendations than niche content viewers. To improve recommendation quality across diverse user groups, we explore three Distributionally Robust Optimization(DRO) methods: Group DRO, Streaming DRO, and Conditional Value at Risk (CVaR) DRO. While Group and Streaming DRO rely on group annotations and struggle with users belonging to multiple groups, CVaR does not require such annotations and can naturally handle overlapping groups. In experiments on two real-world datasets, we show that the DRO methods outperform standard training, with CVaR delivering the best results. Additionally, we find that Group and Streaming DRO are sensitive to the choice of group used for loss computation. Our contributions include (i) a novel application of CVaR to recommenders, (ii) showing that the DRO methods improve group metrics as well as overall performance, and (iii) demonstrating CVaR's effectiveness in the practical scenario of intersecting user groups.