Enhancing Fairness in Meta-learned User Modeling via Adaptive Sampling
Keywords: meta-learning, fairness, user modeling, adaptive sampling
Abstract: Meta-learning has been widely employed to tackle the cold-start problem in user modeling. The core idea is to learn the globally shared meta-initialization parameters for all users and rapidly adapt them to user-specific local parameters. Similar to a guidebook for a new traveler, meta-learning significantly affects decision-making for new users in crucial scenarios, such as career recommendations. Consequently, the issue of fairness in meta-learning has gained paramount importance. Several methods have been proposed to mitigate unfairness in meta-learning and have shown promising results. However, a fundamental question remains unexplored: What is the critical factor leading to unfairness in meta-learned user modeling? Through the theoretical analysis that integrates the meta-learning paradigm with group fairness metrics, we identify group proportion imbalance as a critical factor. Subsequently, in order to mitigate the impact of this factor, we introduce a novel \underline{F}airness-aware \underline{A}daptive \underline{S}ampling framework for me\underline{T}a-learning, abbreviated as FAST. Its core concept involves adaptively adjusting the sampling distribution for different user groups
during the interleaved training process of meta-learning. Furthermore, we provide theoretical guarantees demonstrating the convergence of FAST, showcasing its potential to effectively eliminate unfairness. Finally, empirical experiments conducted on three datasets reveal that FAST effectively enhances fairness while maintaining high accuracy.
Track: User Modeling and Recommendation
Submission Guidelines Scope: Yes
Submission Guidelines Blind: Yes
Submission Guidelines Format: Yes
Submission Guidelines Limit: Yes
Submission Guidelines Authorship: Yes
Student Author: Yes
Submission Number: 327
Loading