Abstract:A well-developed recommendation system can not only leverage multi-typed interactions (such as page view, add-to-cart, and purchase) to better identify user preferences, but also demonstrate high performance, low complexity, and strong interpretability. However, many existing solutions for multi-behavior recommendation fall short of intuitive modeling of real-world scenarios, leading to overly complex models with massive parameters and cumbersome components. In particular, they share two critical limitations: (1) Some pioneering models are built upon the strict assumption of cascade effects across behaviors, which contradicts multifarious behavior paths in practical applications. (2) Existing approaches fail to explicitly capture the unique idiosyncrasies of users and even neglect the inherent nature of items involved in the multi-behavior interactions. To this end, we propose a novel Directed Acyclic Graph Convolutional Network (DA-GCN) for the multi-behavior recommendation task. Specifically, we pinpoint the partial order relations within the monotonic behavior chain and extend it to personalized directed acyclic behavior graphs to exploit behavior dependencies. Then, a GCN-based directed edge encoder is employed to distill rich collaborative signals embodied by each directed edge. In light of the information flows over the directed acyclic structure, we propose an attentive aggregation module to gather messages from all potential antecedent behaviors, representing distinct perspectives to understand the terminated behavior. Thus, we obtain comprehensive representations for the follow-up behavior through learnable distributions over its preceding behaviors, explicitly reflecting personalized interactive patterns of users and underlying properties of items simultaneously. Finally, we design a customized multi-task learning objective for flexible joint optimization. Extensive experiments on public benchmarking datasets fully demonstrate the superiority of DA-GCN with significant performance improvement and computational efficiency over a wide range of state-of-the-art methods. Our code is available at https://github.com/xizhu1022/DA-GCN.

AttentionRank+: A Graph-Based Recommendation Combining Attention Relationship and Multi-Behaviors

ATRank: An Attention-Based User Behavior Modeling Framework for Recommendation

A Graph Neural Network Social Recommendation Algorithm Integrating the Multi-Head Attention Mechanism

Impression-Informed Multi-Behavior Recommender System: A Hierarchical Graph Attention Approach

Multi-Behavior Enhanced Heterogeneous Graph Convolutional Networks Recommendation Algorithm based on Feature-Interaction

Behavior Pattern Mining-based Multi-Behavior Recommendation

Multi-Behavior Recommendation with Personalized Directed Acyclic Behavior Graphs

Multi-Behavior Graph Neural Networks for Recommender System

Self-Attention Mechanism Enhanced User Interests Modeling for Personalized Recommendation Services in Cyber-Physical-Social Systems.

Long- and short-term collaborative attention networks for sequential recommendation

MutualRec: Joint friend and item recommendations with mutualistic attentional graph neural networks

Graph-enhanced and collaborative attention networks for session-based recommendation

Recommendation Based on Users’ Long-Term and Short-Term Interests with Attention

Multi-Head Attentive Social Recommendation

Social Recommendation Based on Users’ Attention and Preference

Multi-behavior Recommendation with Graph Convolutional Networks

A Novel Neighborhood-Augmented Graph Attention Network for Sequential Recommendation.

The Graph-Based Behavior-Aware Recommendation for Interactive News

Matrix Factorization Recommendation Algorithm Based on Attention Interaction

Learning Self-Modulating Attention In Continuous Time Space With Applications To Sequential Recommendation

Hybrid Graph Neural Network Recommendation Based on Multi-Behavior Interaction and Time Sequence Awareness