GraphFM: Graph Factorization Machines for Feature Interaction Modeling
Preprint
Abstract
Factorization machine (FM) is a prevalent approach to modeling pairwise (second-order) feature interactions when dealing with high-dimensional sparse data. However, on the one hand, FM fails to capture higher-order feature interactions suffering from combinatorial expansion, on the other hand, taking into account interaction between every pair of features may introduce noise and degrade prediction accuracy. To solve the problems, we propose a novel approach Graph Factorization Machine (GraphFM) by naturally representing features in the graph structure. In particular, a novel mechanism is designed to select the beneficial feature interactions and formulate them as edges between features. Then our proposed model which integrates the interaction function of FM into the feature aggregation strategy of Graph Neural Network (GNN), can model arbitrary-order feature interactions on the graph-structured features by stacking layers. Experimental results on several real-world datasets has demonstrated the rationality and effectiveness of our proposed approach.Keywords
All Related Versions
This publication has 22 references indexed in Scilit:
- Differentiable Graph Module (DGM) for Graph Convolutional NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2022
- HoAFM: A High-order Attentive Factorization Machine for CTR PredictionInformation Processing & Management, 2020
- Situation Recognition with Graph Neural Networks2017
- 3D Graph Neural Networks for RGBD Semantic Segmentation2017
- Deep & Cross Network for Ad Click Predictions2017
- Neural Factorization Machines for Sparse Predictive Analytics2017
- The More You Know: Using Knowledge Graphs for Image Classification2017
- Product-Based Neural Networks for User Response Prediction2016
- Field-aware Factorization Machines for CTR Prediction2016
- A new model for learning in graph domains2006