Temporal Heterogeneous Interaction Graph Embedding for Next-Item Recommendation

• Published in: Machine Learning and Knowledge Discovery in Databases Vol. 12459; pp. 314 - 329
• Main Authors: Ji, Yugang, Yin, MingYang, Fang, Yuan, Yang, Hongxia, Wang, Xiangwei, Jia, Tianrui, Shi, Chuan
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Long-term habits; Next-item recommendation; Short-term demands; Temporal heterogeneous interaction graph
• ISBN: 3030676633; 9783030676636
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-67664-3_19

In the scenario of next-item recommendation, previous methods attempt to model user preferences by capturing the evolution of sequential interactions. However, their sequential expression is often limited, without modeling complex dynamics that short-term demands can often be influenced by long-term habits. Moreover, few of them take into account the heterogeneous types of interaction between users and items. In this paper, we model such complex data as a Temporal Heterogeneous Interaction Graph (THIG) and learn both user and item embeddings on THIGs to address next-item recommendation. The main challenges involve two aspects: the complex dynamics and rich heterogeneity of interactions. We propose THIG Embedding (THIGE) which models the complex dynamics so that evolving short-term demands are guided by long-term historical habits, and leverages the rich heterogeneity to express the latent relevance of different-typed preferences. Extensive experiments on real-world datasets demonstrate that THIGE consistently outperforms the state-of-the-art methods.