Jointly modeling intra- and inter-transaction dependencies with hierarchical attentive transaction embeddings for next-item recommendation

Shoujin Wang, Longbing Cao, Liang Hu, Shlomo Berkovsky, Xiaoshui Huang, Lin Xiao, Wenpeng Lu

Research output: Contribution to journalArticle

Abstract

A transaction-based recommender system (TBRS) attempts to predict the next item by modeling dependencies in transactional data. Generally, two kinds of dependency considered are intra-transaction dependency and inter-transaction dependency. Most existing TBRSs recommend next item by only modeling the intra-transaction dependency within the current transaction while ignoring inter-transaction dependency with recent transactions that may also affect the next item. However, not all recent transactions are relevant to the current one and next items, such that the relevant ones should be prioritized. In this paper, we propose a novel hierarchical attentive transaction embedding (HATE) model to tackle these issues. Specifically, a two-level attention mechanism integrates both item embeddings and transaction embeddings to build an attentive context representation incorporating both intra- and inter-transaction dependencies and to recommend the next item. Experimental evaluations using two real-world datasets of shopping transactions show that HATE significantly outperforms the state-of-the-art methods in terms of recommendation accuracy.

Original languageEnglish
JournalIEEE Intelligent Systems
Early online date25 May 2020
DOIs
Publication statusE-pub ahead of print - 25 May 2020

Keywords

  • Context modeling
  • Dairy products
  • Data models
  • dependency modelling
  • embedding
  • Intelligent systems
  • Markov processes
  • Predictive models
  • recommendations
  • Recommender systems

Fingerprint Dive into the research topics of 'Jointly modeling intra- and inter-transaction dependencies with hierarchical attentive transaction embeddings for next-item recommendation'. Together they form a unique fingerprint.

Cite this