Dynamic Multi-Behavior Sequence Modeling for Next Item Recommendation

TL;DR

This paper introduces DyMuS and DyMuS+ models for next item recommendation that effectively capture complex user interests from multi-behavior sequences using dynamic routing, outperforming existing methods.

Contribution

The paper proposes novel dynamic multi-behavior sequence models DyMuS and DyMuS+ that better encode user interests by considering sequence characteristics and correlations, with a new dataset for evaluation.

Findings

DyMuS and DyMuS+ outperform baseline models in experiments.

Dynamic routing effectively captures inter-sequence correlations.

DyMuS+ encodes item-level correlations for improved accuracy.

Abstract

Sequential Recommender Systems (SRSs) aim to predict the next item that users will consume, by modeling the user interests within their item sequences. While most existing SRSs focus on a single type of user behavior, only a few pay attention to multi-behavior sequences, although they are very common in real-world scenarios. It is challenging to effectively capture the user interests within multi-behavior sequences, because the information about user interests is entangled throughout the sequences in complex relationships. To this end, we first address the characteristics of multi-behavior sequences that should be considered in SRSs, and then propose novel methods for Dynamic Multi-behavior Sequence modeling named DyMuS, which is a light version, and DyMuS+, which is an improved version, considering the characteristics. DyMuS first encodes each behavior sequence independently, and then combines the encoded sequences using dynamic routing, which dynamically integrates information required in the final result from among many candidates, based on correlations between the sequences. DyMuS+, furthermore, applies the dynamic routing even to encoding each behavior sequence to further capture the correlations at item-level. Moreover, we release a new, large and up-to-date dataset for multi-behavior recommendation. Our experiments on DyMuS and DyMuS+ show their superiority and the significance of capturing the characteristics of multi-behavior sequences.