HEC-GCN: Hypergraph Enhanced Cascading Graph Convolution Network for Multi-Behavior Recommendation

TL;DR

This work tackles sparsity in multi-behavior recommendation by introducing HEC-GCN, which combines a global graph with behavior-specific interaction graphs and learnable hypergraphs in a cascading architecture. A behavior consistency-guided contrastive learning framework enforces both intra- and inter-behavior alignment, enhancing robustness to sparse data. Empirical results on Beibei, Taobao, and Tmall show significant gains over state-of-the-art baselines, supported by comprehensive ablations and analyses. The approach offers a practical pathway to more accurate recommendations in systems with rich, multi-behavior user interactions.

Abstract

Multi-behavior recommendation (MBR) has garnered growing attention recently due to its ability to mitigate the sparsity issue by inferring user preferences from various auxiliary behaviors to improve predictions for the target behavior. Although existing research on MBR has yielded impressive results, they still face two major limitations. First, previous methods mainly focus on modeling fine-grained interaction information between users and items under each behavior, which may suffer from sparsity issue. Second, existing models usually concentrate on exploiting dependencies between two consecutive behaviors, leaving intra- and inter-behavior consistency largely unexplored. To the end, we propose a novel approach named Hypergraph Enhanced Cascading Graph Convolution Network for multi-behavior recommendation (HEC-GCN). To be specific, we first explore both fine- and coarse-grained correlations among users or items of each behavior by simultaneously modeling the behavior-specific interaction graph and its corresponding hypergraph in a cascaded manner. Then, we propose a behavior consistency-guided alignment strategy that ensures consistent representations between the interaction graph and its associated hypergraph for each behavior, while also maintaining representation consistency across different behaviors. Extensive experiments and analyses on three public benchmark datasets demonstrate that our proposed approach is consistently superior to previous state-of-the-art methods due to its capability to effectively attenuate the sparsity issue as well as preserve both intra- and inter-behavior consistencies. The code is available at https://github.com/marqu22/HEC-GCN.git.

Figures (7)

• Figure 1: The overall architecture of HEC-GCN. We utilize three behaviors (i.e., view, cart and buy) as an example, wherein buy is the target behavior.
• Figure 2: Model performance with respect to different interaction density degrees.
• Figure 3: The impact of different cold-start ratios on model performance.
• Figure 4: The Impact of auxiliary behaviors on performance.
• Figure 5: Impact of behavior sequence order on performance in the cascading architecture.