DemiNet: Dependency-Aware Multi-Interest Network with Self-Supervised Graph Learning for Click-Through Rate Prediction

TL;DR

A novel model named DemiNet (short for DEpendency-Aware Multi-Interest Network) is proposed to address the above two issues and significantly improves the overall recommendation performance over several state-of-the-art baselines.

Abstract

In this paper, we propose a novel model named DemiNet (short for DEpendency-Aware Multi-Interest Network) to address the above two issues. To be specific, we first consider various dependency types between item nodes and perform dependency-aware heterogeneous attention for denoising and obtaining accurate sequence item representations. Secondly, for multiple interests extraction, multi-head attention is conducted on top of the graph embedding. To filter out noisy inter-item correlations and enhance the robustness of extracted interests, self-supervised interest learning is introduced to the above two steps. Thirdly, to aggregate the multiple interests, interest experts corresponding to different interest routes give rating scores respectively, while a specialized network assigns the confidence of each score. Experimental results on three real-world datasets demonstrate that the proposed DemiNet significantly improves the overall recommendation performance over several state-of-the-art baselines. Further studies verify the efficacy and interpretability benefits brought by the fine-grained user interest modeling.

Figures (5)

• Figure 1: A user usually has multiple interests when browsing an e-commerce platform. His current behavior is highly correlated with short-term contextual and long-term similarity dependencies.
• Figure 2: Illustration of the DemiNet model, which is made up of three modules and a self-supervised interest learning task.
• Figure 3: Hyper-Parameter Study
• Figure 4: Multiple Interests Visualization using t-SNE
• Figure 5: Case Study Of the Assigned Confidence