LLM-Enhanced Multimodal Fusion for Cross-Domain Sequential Recommendation
This work addresses the problem of predicting user behavior across multiple domains for e-commerce applications, representing an incremental improvement through multimodal data fusion.
The paper tackles cross-domain sequential recommendation by proposing LLM-Enhanced Multimodal Fusion (LLM-EMF), which integrates visual and textual data using LLMs and CLIP to improve item representations, resulting in consistent performance gains over existing methods on four e-commerce datasets.
Cross-Domain Sequential Recommendation (CDSR) predicts user behavior by leveraging historical interactions across multiple domains, focusing on modeling cross-domain preferences and capturing both intra- and inter-sequence item relationships. We propose LLM-Enhanced Multimodal Fusion for Cross-Domain Sequential Recommendation (LLM-EMF), a novel and advanced approach that enhances textual information with Large Language Models (LLM) knowledge and significantly improves recommendation performance through the fusion of visual and textual data. Using the frozen CLIP model, we generate image and text embeddings, thereby enriching item representations with multimodal data. A multiple attention mechanism jointly learns both single-domain and cross-domain preferences, effectively capturing and understanding complex user interests across diverse domains. Evaluations conducted on four e-commerce datasets demonstrate that LLM-EMF consistently outperforms existing methods in modeling cross-domain user preferences, thereby highlighting the effectiveness of multimodal data integration and its advantages in enhancing sequential recommendation systems. Our source code will be released.