DiffCL: A Diffusion-Based Contrastive Learning Framework with Semantic Alignment for Multimodal Recommendations

TL;DR

DiffCL is a novel diffusion-based contrastive learning framework that enhances multimodal recommendation accuracy by reducing noise, aligning semantic information across modalities, and addressing data sparsity through graph-based features.

Contribution

The paper introduces a diffusion model for contrastive views, semantic alignment via stable ID embeddings, and an item-item graph to improve multimodal recommendation systems.

Findings

DiffCL outperforms existing methods on three public datasets.

The framework effectively reduces noise impact during contrastive learning.

Semantic alignment improves cross-modal consistency and recommendation accuracy.

Abstract

Multimodal recommendation systems integrate diverse multimodal information into the feature representations of both items and users, thereby enabling a more comprehensive modeling of user preferences. However, existing methods are hindered by data sparsity and the inherent noise within multimodal data, which impedes the accurate capture of users' interest preferences. Additionally, discrepancies in the semantic representations of items across different modalities can adversely impact the prediction accuracy of recommendation models. To address these challenges, we introduce a novel diffusion-based contrastive learning framework (DiffCL) for multimodal recommendation. DiffCL employs a diffusion model to generate contrastive views that effectively mitigate the impact of noise during the contrastive learning phase. Furthermore, it improves semantic consistency across modalities by aligning distinct visual and textual semantic information through stable ID embeddings. Finally, the introduction of the Item-Item Graph enhances multimodal feature representations, thereby alleviating the adverse effects of data sparsity on the overall system performance. We conduct extensive experiments on three public datasets, and the results demonstrate the superiority and effectiveness of the DiffCL.