Multi-Interactive Attention Network for Fine-grained Feature Learning in CTR Prediction

TL;DR

This paper introduces MIAN, a novel neural network model that captures complex, fine-grained user preferences and feature interactions to improve CTR prediction accuracy in recommendation systems.

Contribution

The paper proposes a Multi-Interactive Attention Network (MIAN) that effectively models multiple feature interactions and user preferences, addressing limitations of existing sequential methods.

Findings

MIAN outperforms baseline models on three datasets.

Online A/B testing shows significant CTR improvement.

The model effectively captures long-term and fine-grained user preferences.

Abstract

In the Click-Through Rate (CTR) prediction scenario, user's sequential behaviors are well utilized to capture the user interest in the recent literature. However, despite being extensively studied, these sequential methods still suffer from three limitations. First, existing methods mostly utilize attention on the behavior of users, which is not always suitable for CTR prediction, because users often click on new products that are irrelevant to any historical behaviors. Second, in the real scenario, there exist numerous users that have operations a long time ago, but turn relatively inactive in recent times. Thus, it is hard to precisely capture user's current preferences through early behaviors. Third, multiple representations of user's historical behaviors in different feature subspaces are largely ignored. To remedy these issues, we propose a Multi-Interactive Attention Network (MIAN) to comprehensively extract the latent relationship among all kinds of fine-grained features (e.g., gender, age and occupation in user-profile). Specifically, MIAN contains a Multi-Interactive Layer (MIL) that integrates three local interaction modules to capture multiple representations of user preference through sequential behaviors and simultaneously utilize the fine-grained user-specific as well as context information. In addition, we design a Global Interaction Module (GIM) to learn the high-order interactions and balance the different impacts of multiple features. Finally, Offline experiment results from three datasets, together with an Online A/B test in a large-scale recommendation system, demonstrate the effectiveness of our proposed approach.