Deep Session Heterogeneity-aware Network for Click Through Rate Prediction

Xin Zhang,Zengmao Wang,Bo Du,Jia Wu,Xiao Zhang,Erli Meng

DOI: https://doi.org/10.1109/tkde.2024.3421594

IF: 9.235

2024-01-01

IEEE Transactions on Knowledge and Data Engineering

Abstract:CTR (Click-Through Rate) prediction plays an essential role in online advertising systems. Most existing works attempt to capture users' interests from sessions by assuming that behaviors within a session are homogeneous. However, user interest may change frequently. Thus it is hard to guarantee that behaviors in a session are homogeneous, resulting in users' interests extracted from sessions being biased. In this paper, we propose a model named Deep Session Heterogeneity-aware Network (DSHN) by learning the relationships of behaviors within sessions and the relevance between the session and target item to alleviate the influence of irrelevant or heterogeneous sessions. We design a heterogeneity-aware mechanism to learn the heterogeneity of items within a session. Then we further design two modules: the Session Heterogeneity Learning module and the Relevance Inference module. The Session Heterogeneity Learning module weighs each session by summarizing the variation of session interest with and without any behavior. The relevance Inference module learns the relevance between the target item and each session in a similar way by learning session interest with and without the target item. Extensive experiments on four datasets demonstrate that our proposed DSHN achieves better results compared to the state-of-the-art.