Abstract:Modeling feature interactions is of crucial importance to predict click-through rate (CTR) in industrial recommender systems. However, manually crafting cross features usually requires extensive domain knowledge and labor-intensive feature engineering to obtain the desired cross features. To alleviate this problem, the factorization machine (FM) is proposed to model feature interactions from raw features automatically. In particular, it embeds each feature in a vector representation and discovers second-order interactions as the product of two feature representations. In order to learn nonlinear and complex patterns, recent works, such as NFM, PIN, and DeepFM, exploited deep learning techniques to capture higher-order feature interactions. These approaches lack guarantees about the effectiveness of high-order pattern as they model feature interactions in a rather implicit way. To address this limitation, xDeepFM is recently proposed to generate high-order interactions of features in an explicit fashion, where multiple interaction networks are stacked. Nevertheless, xDeepFM suffers from rather high complexity which easily leads to overfitting.In this paper, we develop a more expressive but lightweight solution based on FM, named High-order Attentive Factorization Machine (HoAFM), by accounting for the higher-order sparse feature interactions in an explicit manner. Beyond the linearity of FM, we devise a cross interaction layer, which updates a feature’s representation by aggregating the representations of other co-occurred features. In addition, we perform a bit-wise attention mechanism to determine the different importance of co-occurred features on the granularity of dimensions. By stacking multiple cross interaction layers, we can inject high-order feature interactions into feature representation learning, in order to establish expressive and informative cross features. Extensive experiments are performed on two benchmark datasets, Criteo and Avazu, to demonstrate the rationality and effectiveness of HoAFM. Empirical results suggest that HoAFM achieves significant improvement over other state-of-the-art methods, such as NFM and xDeepFM. https://github.com/zltao/HoAFM.

A Dual Adaptive Factorization Network for CTR Prediction.

Click-Through Rate Prediction Algorithm Based on Modeling of Implicit High-Order Feature Importance

Towards Deeper, Lighter and Interpretable Cross Network for CTR Prediction

Input Enhanced Logarithmic Factorization Network for CTR Prediction

Field-Aware Neural Factorization Machine for Click-Through Rate Prediction

A Dual Input-aware Factorization Machine for CTR Prediction

DeepFM: A Factorization-Machine based Neural Network for CTR Prediction

DCNv3: Towards Next Generation Deep Cross Network for CTR Prediction

HoAFM: A High-order Attentive Factorization Machine for CTR Prediction

A Dual Adaptive Interaction Click-Through Rate Prediction Based on Attention Logarithmic Interaction Network

Deep & Attentional Crossing Network for Click-through Rate Prediction

TF4CTR: Twin Focus Framework for CTR Prediction via Adaptive Sample Differentiation

Feature Interaction Fusion Self-Distillation Network For CTR Prediction

FAT-DeepFFM: Field Attentive Deep Field-aware Factorization Machine

Coarse-to-fine: A dual-view attention network for click-through rate prediction

FINAL:Factorized Interaction Layer for CTR Prediction

An Efficient Deep Interaction Network for Click-Through Rate Prediction.

Context-aware cross feature attentive network for click-through rate predictions

Feature Generation by Convolutional Neural Network for Click-Through Rate Prediction

MCRF: Enhancing CTR Prediction Models Via Multi-channel Feature Refinement Framework

FLEN: Leveraging Field for Scalable CTR Prediction