From Overlooked to Explored: Recovering Item Relations via Mixture of Perspectives for Sequential Recommendation

2026-08-12

AI Summary

This work addresses a critical limitation in existing self-attention–based sequential recommendation models, which suffer from similarity bias—overemphasizing interactions among homogeneous items while neglecting heterogeneous associations that reflect nuanced user preferences, thereby constraining performance. The study is the first to formally identify and characterize this bias and introduces PRISM, a novel module integrated into the Transformer architecture. PRISM employs a multi-view mechanism that jointly models both homogeneous and heterogeneous item relationships: an affinity view reinforces intra-category similarities, while a contrastive view uncovers underrepresented cross-category dependencies. Extensive experiments across seven real-world datasets demonstrate that PRISM consistently outperforms state-of-the-art baselines, significantly enhancing both recommendation accuracy and model generalization.

Abstract

Capturing user preference from a user's interaction sequence is the central challenge of Sequential Recommendation (SR). This preference intuitively emerges from inter-item relations: each item transition reflects a preference embedded in the relations between items, making the faithful capture of these relations essential for accurate recommendation. For this reason, self-attention is dominant in sequential recommendation for its ability to compute pairwise item interactions, yet our empirical analysis reveals that it consistently suffers from similarity bias across various types of transformer-based SR models: dot-product attention scores disproportionately favor similar items, systematically overlooking heterogeneous relations with meaningful preference signals and directly limiting recommendation performance. To address this, we propose PRISM (Perspective-based Relational Insight Synthesis Module), a module that re-examines item relations from multiple perspectives. PRISM employs K Perspective Lenses to calibrate attention from distinct viewpoints, combining an Affinity View that refines homogeneous relations and a Contrast View that exposes heterogeneous ones suppressed by similarity bias, enabling the model to capture the full spectrum of user preferences. Extensive experiments on seven real-world benchmarks demonstrate that PRISM consistently outperforms state-of-the-art baselines. Our code is available at https://github.com/327aem/PRISM/.

Problem

• sequential recommendation
• item relations
• similarity bias
• self-attention
• heterogeneous relations

Innovation

• similarity bias
• multi-perspective attention
• heterogeneous item relations
• sequential recommendation
• PRISM