Deep Multi-Representational Item Network for CTR Prediction

Click-through rate (CTR) prediction is essential in the modelling of a recommender system. Previous studies mainly focus on user behavior modelling, while few of them consider candidate item representations. This makes the models strongly dependent on user representations, and less effective when user behavior is sparse. Furthermore, most existing works regard the candidate item as one fixed embedding and ignore the multi-representational characteristics of the item. To handle the above issues, we propose a Deep multi-Representational Item NetworK (DRINK) for CTR prediction. Specifically, to tackle the sparse user behavior problem, we construct a sequence of interacting users and timestamps to represent the candidate item; to dynamically capture the characteristics of the item, we propose a transformer-based multi-representational item network consisting of a multi-CLS representation submodule and contextualized global item representation submodule. In addition, we propose to decouple the time information and item behavior to avoid information overwhelming. Outputs of the above components are concatenated and fed into a MLP layer to fit the CTR. We conduct extensive experiments on real-world datasets of Amazon and the results demonstrate the effectiveness of the proposed model.

Automated summary

CTR prediction plays a crucial role in modeling recommender systems. Previous studies have mainly focused on user behavior modeling, neglecting the representations of candidate items. In this paper, we propose a Deep multi-Representational Item NetworK (DRINK) that addresses the sparse user behavior problem and captures the multi-representational characteristics of candidate items using a transformer-based approach. We also decouple time information and item behavior to avoid information overwhelming. Experimental results on real-world datasets demonstrate the effectiveness of the proposed model.