Disentangled Self-Attentive Neural Networks for Click-Through Rate Prediction

TL;DR

This paper introduces DESTINE, a novel neural network framework that explicitly separates feature importance and pairwise interactions to improve click-through rate prediction accuracy while maintaining efficiency.

Contribution

The paper proposes DESTINE, a disentangled self-attentive neural network that explicitly models unary feature importance separately from pairwise interactions for CTR prediction.

Findings

DESTINE outperforms state-of-the-art models on benchmark datasets.

It maintains computational efficiency despite increased modeling complexity.

Extensive experiments validate the effectiveness of the disentangled approach.

Abstract

Click-Through Rate (CTR) prediction, whose aim is to predict the probability of whether a user will click on an item, is an essential task for many online applications. Due to the nature of data sparsity and high dimensionality of CTR prediction, a key to making effective prediction is to model high-order feature interaction. An efficient way to do this is to perform inner product of feature embeddings with self-attentive neural networks. To better model complex feature interaction, in this paper we propose a novel DisentanglEd Self-atTentIve NEtwork (DESTINE) framework for CTR prediction that explicitly decouples the computation of unary feature importance from pairwise interaction. Specifically, the unary term models the general importance of one feature on all other features, whereas the pairwise interaction term contributes to learning the pure impact for each feature pair. We conduct extensive experiments using two real-world benchmark datasets. The results show that DESTINE not only maintains computational efficiency but achieves consistent improvements over state-of-the-art baselines.