Propagating Similarity, Mitigating Uncertainty: Similarity Propagation-enhanced Uncertainty for Multimodal Recommendation

TL;DR

This paper introduces SPUMR, a framework that models and reduces uncertainty in multimodal recommendation systems by leveraging similarity graphs and adaptive feature fusion, leading to improved performance.

Contribution

The paper proposes a novel similarity propagation framework that explicitly models and mitigates modality-specific uncertainty in multimodal recommendation systems.

Findings

SPUMR outperforms existing methods on three benchmark datasets.

The similarity graphs effectively refine feature representations.

Adaptive fusion improves recommendation accuracy.

Abstract

Multimodal Recommendation (MMR) systems are crucial for modern platforms but are often hampered by inherent noise and uncertainty in modal features, such as blurry images, diverse visual appearances, or ambiguous text. Existing methods often overlook this modality-specific uncertainty, leading to ineffective feature fusion. Furthermore, they fail to leverage rich similarity patterns among users and items to refine representations and their corresponding uncertainty estimates. To address these challenges, we propose a novel framework, Similarity Propagation-enhanced Uncertainty for Multimodal Recommendation (SPUMR). SPUMR explicitly models and mitigates uncertainty by first constructing the Modality Similarity Graph and the Collaborative Similarity Graph to refine representations from both content and behavioral perspectives. The Uncertainty-aware Preference Aggregation module then adaptively fuses the refined multimodal features, assigning greater weight to more reliable modalities. Extensive experiments on three benchmark datasets demonstrate that SPUMR achieves significant improvements over existing leading methods.