Exploiting Repeated Behavior Pattern and Long-term Item dependency for Session-based Recommendation

Session-based recommendation (SBR) is a challenging task, which aims to predict users' future interests based on anonymous behavior sequences. Existing methods for SBR leverage powerful representation learning approaches to encode sessions into a low dimensional space. However, all the existing studies focus on the item transitions in the session, without modeling the behavior patterns, which are strong clues to capture the preference of users. Further, the long-term dependency within the session is neglected in most of the current methods. To this end, we propose a novel Repeat-aware Neural Mechanism for Session-based Recommendation (RNMSR). Specifically, we introduce repeated behavior pattern into SBR, which contains the potential intent information and critical item frequency signal. Furthermore, we also built a similarity-based session graph based on long-term dependencies within a session. Extensive experiments conducted on two benchmark E-commerce datasets, Yoochoose and Diginetica demonstrate our proposed method outperforms the state-of-the-art methods consistently.