Graph-Structured Driven Dual Adaptation for Mitigating Popularity Bias

2025-03-30

AI Summary

To address representation collapse and degraded generalization caused by popularity bias in recommender systems, this paper proposes a graph-structure-driven dual-adaptation framework. Methodologically, it introduces a novel hierarchical adaptive alignment mechanism to overcome the oversmoothing bottleneck inherent in supervised alignment within deep graph convolutional networks (GCNs); designs a dynamic contrastive weighting strategy based on real-time Gini coefficient estimation, eliminating reliance on fixed hyperparameters; and incorporates Frobenius norm-based layerwise adaptive decay of the adjacency matrix alongside conditional entropy theory to jointly model graph structural properties and item distribution characteristics. Evaluated on three benchmark datasets, the proposed method significantly mitigates popularity bias, consistently outperforms state-of-the-art models, and notably improves exposure and click-through rates for long-tail items.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Recommender Systems
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Popularity bias challenges recommender systems by causing uneven recommendation performance and amplifying the Matthew effect. Limited user-item interactions confine unpopular items within embedding neighborhoods of few users, leading to representation collapse and reduced model generalization. Existing supervised alignment and reweighting methods mitigate this bias but have key limitations: (1) ignoring inherent variability across Graph Convolutional Networks (GCNs) layers, causing negative effects in deeper layers; (2) reliance on fixed hyperparameters to balance item popularity, restricting adaptability and increasing complexity. To address these issues, we propose the Graph-Structured Dual Adaptation Framework (GSDA). Our theoretical analysis identifies a crucial limitation of supervised alignment methods caused by over-smoothing in GCNs. As GCN layers deepen, popular and unpopular items increasingly lose distinctiveness, quantified by reduced conditional entropy. This diminished distinctiveness weakens supervised alignment effectiveness in mitigating popularity bias. Motivated by this, GSDA captures structural and distribution characteristics from the adjacency matrix through a dual adaptive strategy. First, a hierarchical adaptive alignment mechanism uses the adjacency matrix's Frobenius norm for layer-specific weight decay, countering conditional entropy reduction effects at deeper layers. Second, a distribution-aware dynamic contrast weighting strategy, guided by a real-time Gini coefficient, removes dependence on fixed hyperparameters, enabling adaptability to diverse data. Experiments on three benchmark datasets demonstrate GSDA significantly alleviates popularity bias and consistently outperforms state-of-the-art recommendation methods.

Problem

• Mitigating popularity bias in recommender systems
• Addressing representation collapse in GCNs for unpopular items
• Overcoming limitations of fixed hyperparameters in bias mitigation

Innovation

• Hierarchical adaptive alignment with Frobenius norm
• Dynamic contrast weighting using Gini coefficient
• Graph-structured dual adaptation for bias mitigation