Sequential Recommendation Models: A Graph-based Perspective

Abstract

Recommender systems (RS) traditionally leverage the users' rich interaction data with the system, but ignore the sequential dependency of items. Sequential recommender systems aim to predict the next item the user will interact with (e.g., click on, purchase, or listen to) based on the preceding interactions of the user with the system. Current state-of-the-art approaches focus on transformer-based architectures and graph neural networks. Specifically, graph-based modeling of sequences has been shown to be state-of-the-art by introducing a structured, inductive bias into the recommendation learning framework. In this work, we outline our research into designing novel graph-based methods for sequential recommendation.