Rethinking Item Importance in Session-based Recommendation
This work addresses session-based recommendation for users with anonymous sessions, offering an incremental improvement over existing methods.
The paper tackles the problem of session-based recommendation by proposing SR-IEM, which estimates item importance to better capture user intent, resulting in improved Recall and MRR on benchmark datasets with reduced computational complexity.
Session-based recommendation aims to predict users' based on anonymous sessions. Previous work mainly focuses on the transition relationship between items during an ongoing session. They generally fail to pay enough attention to the importance of the items in terms of their relevance to user's main intent. In this paper, we propose a Session-based Recommendation approach with an Importance Extraction Module, i.e., SR-IEM, that considers both a user's long-term and recent behavior in an ongoing session. We employ a modified self-attention mechanism to estimate item importance in a session, which is then used to predict user's long-term preference. Item recommendations are produced by combining the user's long-term preference and current interest as conveyed by the last interacted item. Experiments conducted on two benchmark datasets validate that SR-IEM outperforms the start-of-the-art in terms of Recall and MRR and has a reduced computational complexity.