Multi-granularity Item-based Contrastive Recommendation

TL;DR

This paper introduces MicRec, a multi-granularity contrastive learning framework for recommendation that leverages item-related information at feature, semantic, and session levels to improve matching accuracy.

Contribution

It proposes a novel multi-aspect item-based contrastive learning framework with three auxiliary tasks, enhancing item representation learning in recommendation systems.

Findings

Effective in offline and online evaluations

Improves recommendation accuracy across multiple datasets

Deployed in a real-world system affecting millions

Abstract

Contrastive learning (CL) has shown its power in recommendation. However, most CL-based recommendation models build their CL tasks merely focusing on the user's aspects, ignoring the rich diverse information in items. In this work, we propose a novel Multi-granularity item-based contrastive learning (MicRec) framework for the matching stage (i.e., candidate generation) in recommendation, which systematically introduces multi-aspect item-related information to representation learning with CL. Specifically, we build three item-based CL tasks as a set of plug-and-play auxiliary objectives to capture item correlations in feature, semantic and session levels. The feature-level item CL aims to learn the fine-grained feature-level item correlations via items and their augmentations. The semantic-level item CL focuses on the coarse-grained semantic correlations between semantically related items. The session-level item CL highlights the global behavioral correlations of items from users' sequential behaviors in all sessions. In experiments, we conduct both offline and online evaluations on real-world datasets, verifying the effectiveness and universality of three proposed CL tasks. Currently, MicRec has been deployed on a real-world recommender system, affecting millions of users. The source code will be released in the future.