Learning Dynamic User Behavior Based on Error-driven Event Representation

Abstract

Understanding the evolution of large graphs over time is of significant importance in user behavior understanding and prediction. Modeling user behavior with temporal networks has gained increasing attention in recent years since it allows capturing users’ dynamic preferences and predicting their next actions. Recently, some approaches have been proposed to model user behavior. However, these methods suffer from two problems: they work on static data, which ignores the dynamic evolution, or they model the whole behavior sequences directly by recurrent neural networks and thus suffer from noisy information. To tackle these problems, we propose a dynamic user behavior learning algorithm called LDBR. It views user behaviors as a set of dynamic events and uses recent event embedding to predict future user behavior and infer the current semantic labels. Specifically, we propose a new strategy to automatically learn a good event embedding in behavior sequence by introducing a smooth sampling strategy and minimizing the temporal link prediction error. It is hard to obtain real-world datasets with evolving labels. Thus in this paper, we provide a new dynamic network dataset with evolving labels called Arxiv and make it publicly available. Based on the Arxiv dataset, we conduct a case study to verify the quality of event embedding. Extensive experiments on temporal link prediction tasks further demonstrate the effectiveness of the LDBR model.