Decision-Making Context Interaction Network for Click-Through Rate Prediction

Click-through rate (CTR) prediction is crucial in recommendation and online advertising systems. Existing methods usually model user behaviors, while ignoring the informative context which influences to make a click decision, e.g., pages pre-ranking candidates that inform inferences about interests, leading suboptimal performance. In this paper, we propose Decision-Making Context Interaction Network (DCIN), deploys carefully designed Unit (CIU) learn decision-making contexts thus benefits CTR prediction. addition, relationship between different sources explored by proposed Adaptive Interest Aggregation (AIAU) improve further. experiments on public industrial datasets, DCIN significantly outperforms state-of-the-art methods. Notably, has obtained improvement of CTR+2.9%/CPM+2.1%/GMV+1.5% for A/B testing served main traffic Meituan Waimai system.