Match4Rec: A Novel Recommendation Algorithm Based on Bidirectional Encoder Representation with the Matching Task

TL;DR

Match4Rec introduces a bidirectional encoder with a matching task to better capture both local and global user preferences in sequential recommendation, significantly improving performance over existing models.

Contribution

The paper presents a novel bidirectional recommendation model that combines local behavioral purposes with global preferences using a matching task and a new sample production method.

Findings

Outperforms state-of-the-art models in empirical tests

Effectively captures both local and global user preferences

Enhances user representation learning in sequential data

Abstract

Characterizing users' interests accurately plays a significant role in an effective recommender system. The sequential recommender system can learn powerful hidden representations of users from successive user-item interactions and dynamic users' preferences. To analyze such sequential data, conventional methods mainly include Markov Chains (MCs) and Recurrent Neural Networks (RNNs). Recently, the use of self-attention mechanisms and bi-directional architectures have gained much attention. However, there still exists a major limitation in previous works that they only model the user's main purposes in the behavioral sequences separately and locally, and they lack the global representation of the user's whole sequential behavior. To address this limitation, we propose a novel bidirectional sequential recommendation algorithm that integrates the user's local purposes with the global preference by additive supervision of the matching task. We combine the mask task with the matching task in the training process of the bidirectional encoder. A new sample production method is also introduced to alleviate the effect of mask noise. Our proposed model can not only learn bidirectional semantics from users' behavioral sequences but also explicitly produces user representations to capture user's global preference. Extensive empirical studies demonstrate our approach considerably outperforms various state-of-the-art models.