Basket-Enhanced Heterogenous Hypergraph for Price-Sensitive Next Basket Recommendation
Yuening Zhou, Yulin Wang, Qian Cui, Xinyu Guan, Francisco Cisternas
TL;DR
This work tackles Next Basket Recommendation by incorporating price sensitivity into a basket-centric, multi-feature learning framework. It introduces BDHH, a Basket-augmented Dynamic Heterogeneous Hypergraph that marries a heterogeneous multi-relational graph with a basket-guided augmentation network and price-aware user behavior modeling. Core contributions include a hypergraph with item-ID, item-price, and category features, a unified global hybrid encoder, and a basket-driven mechanism that dynamically refines item representations; together with a training objective based on cross-entropy losses. Experiments on real-world datasets demonstrate substantial gains in NDCG and Hit metrics, with ablations confirming the critical roles of price information and basket-guided augmentation for accurate next-basket predictions.
Abstract
Next Basket Recommendation (NBR) is a new type of recommender system that predicts combinations of items users are likely to purchase together. Existing NBR models often overlook a crucial factor, which is price, and do not fully capture item-basket-user interactions. To address these limitations, we propose a novel method called Basket-augmented Dynamic Heterogeneous Hypergraph (BDHH). BDHH utilizes a heterogeneous multi-relational graph to capture the intricate relationships among item features, with price as a critical factor. Moreover, our approach includes a basket-guided dynamic augmentation network that could dynamically enhances item-basket-user interactions. Experiments on real-world datasets demonstrate that BDHH significantly improves recommendation accuracy, providing a more comprehensive understanding of user behavior.