Multi-aspect Interest Neighbor-augmented Network for Next-basket Recommendation

Next-basket recommendation (NBR) is a type task of recommendation that focuses on mining user preference based on the sequential basket records in which users purchase multiple items at a time. Limited by the sparsity brought by short-term user interaction behaviors, existing methods generally fail to mine fine-grained and complete user interests representations, resulting in weak recommendation performance. To address this problem, we propose a novel Multi-Aspect Interest Neighbor-augmented Network (MINN) to capture fine-grained and complete user preference representations for next basket prediction. Specifically, we first design a multi-aspect interest encoder to learn the representations of users and items in multi-aspect interest spaces. Then, the semantic neighbors are filtered to enhance user’s interest representations by a semantic neighbor augmentation mechanism. Finally, user’s interest representations and user’s repeated purchase behavior are jointly considered for implementing the next basket prediction. Extensive experimental studies on two benchmark datasets demonstrate that MINN outperforms several representative NBR methods and achieves new state-of-the-art results.