Search papers, labs, and topics across Lattice.
This paper introduces FSGR, a fairness optimization framework designed to address the Token Frequency Bias in SID-based generative recommendation systems, where high-frequency tokens are over-predicted and low-frequency tokens are under-predicted. By employing Optimal Transport-based Assignment Optimization and a Dual-Criteria Re-anchor mechanism during SID construction, along with Hierarchical Frequency Calibration in a two-stage training strategy, FSGR significantly enhances fairness in recommendations. Experimental results across three datasets show that FSGR achieves over a 20% improvement in Gini fairness while preserving competitive accuracy in recommendations.
Overcoming Token Frequency Bias could redefine fairness in generative recommendation systems, achieving over 20% improvement in fairness metrics without sacrificing accuracy.
Semantic ID (SID)-based generative recommendation has recently achieved remarkable success. However, existing methods suffer from a previously overlooked fairness issue, which we term \textbf{Token Frequency Bias}, where high-frequency SID tokens are systematically over-predicted while low-frequency SID tokens are under-predicted. This bias originates from the combined effects of imbalanced semantic codebooks during SID construction, and popularity bias together with the maximum likelihood estimation objective during recommendation training, resulting in unfair exposure across item categories. Existing SID methods mainly focus on improving codebook quality and overlook the impact of token frequency imbalance on downstream recommendation fairness, while LLM debiasing methods often yield suboptimal results when directly applied to SID-based recommendation, due to the hierarchical semantics of SID tokens. To address this issue, we propose \textbf{FSGR}, a fairness optimization framework for SID-based generative recommendation. During SID construction, FSGR employs OT-based Assignment Optimization and Dual-Criteria Re-anchor mechanism to form a more balanced SID representation space. During recommendation training, it adopts a two-stage training strategy and introduces Hierarchical Frequency Calibration for layer-specific fairness fine-tuning. Experiments on three public datasets with three backbone models demonstrate that FSGR mitigates token frequency bias and delivers an average Gini fairness improvement of over 20\% while maintaining competitive recommendation accuracy.
