Context-embedded Hypergraph Attention Network And Self-attention For Session Recommendation

The article discusses the challenges of modeling user intention in session recommendation and proposes a novel session-based model called C-HAN that utilizes context-embedded hypergraph attention network and self-attention to capture inherent consistency and sequential dependencies between items. Experimental results show that C-HAN outperforms state-of-the-art methods in precision, recall, and MRR metrics.