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

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 in CTR prediction, a key to making effective prediction is to model high-order feature interaction among feature fields. To explicitly model high-order feature interaction, an efficient way 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 importance from pairwise interaction. Specifically, the unary term models the general impact of one feature on all other features, whereas the whitened pairwise interaction term contributes to learning the pure importance score for each feature interaction. We conduct extensive experiments framework using two real-world benchmark datasets. The results show that DESTINE not only maintains computational efficiency but obtains performance improvements over state-of-the-art baselines.