FiBiNET: Combining Feature Importance and Bilinear feature Interaction for Click-Through Rate Prediction

Abstract

Advertising and feed ranking are essential to many Internet companies such as Facebook and Sina Weibo. Among many real-world advertising and feed ranking systems, click through rate (CTR) prediction plays a central role. There are many proposed models in this field such as logistic regression, tree based models, factorization machine based models and deep learning based CTR models. However, many current works calculate the feature interactions in a simple way such as Hadamard product and inner product and they care less about the importance of features. In this paper, a new model named FiBiNET as an abbreviation for Feature Importance and Bilinear feature Interaction NETwork is proposed to dynamically learn the feature importance and fine-grained feature interactions. On the one hand, the FiBiNET can dynamically learn the importance of features via the Squeeze-Excitation network (SENET) mechanism; on the other hand, it is able to effectively learn the feature interactions via bilinear function. We conduct extensive experiments on two real-world datasets and show that our shallow model outperforms other shallow models such as factorization machine(FM) and field-aware factorization machine(FFM). In order to improve performance further, we combine a classical deep neural network(DNN) component with the shallow model to be a deep model. The deep FiBiNET consistently outperforms the other state-of-the-art deep models such as DeepFM and extreme deep factorization machine(XdeepFM).

arXiv comment8 pages,5 figures

References in corpus (1)

• Weakly Supervised Deep Learning for Thoracic Disease Classification and Localization on Chest X-rays

Cited by in corpus (24)

• Click-Through Rate Prediction in Online Advertising: A Literature Review
• An Embedding Learning Framework for Numerical Features in CTR Prediction
• CL4CTR: A Contrastive Learning Framework for CTR Prediction
• EulerNet: Adaptive Feature Interaction Learning via Euler's Formula for CTR Prediction
• Enhancing CTR Prediction with Context-Aware Feature Representation Learning
• Directed Acyclic Graph Factorization Machines for CTR Prediction via Knowledge Distillation
• Ads Recommendation in a Collapsed and Entangled World
• Split Two-Tower Model for Efficient and Privacy-Preserving Cross-device Federated Recommendation
• A General Method For Automatic Discovery of Powerful Interactions In Click-Through Rate Prediction
• FiBiNet++: Reducing Model Size by Low Rank Feature Interaction Layer for CTR Prediction
• AMEIR: Automatic Behavior Modeling, Interaction Exploration and MLP Investigation in the Recommender System
• Climber: Toward Efficient Scaling Laws for Large Recommendation Models
• RAT: Retrieval-Augmented Transformer for Click-Through Rate Prediction
• AIE: Auction Information Enhanced Framework for CTR Prediction in Online Advertising
• Risk-Aware Bid Optimization for Online Display Advertisement
• DLF: Enhancing Explicit-Implicit Interaction via Dynamic Low-Order-Aware Fusion for CTR Prediction
• LARR: Large Language Model Aided Real-time Scene Recommendation with Semantic Understanding
• Adaptive Learning on User Segmentation: Universal to Specific Representation via Bipartite Neural Interaction
• Toward a benchmark for CTR prediction in online advertising: datasets, evaluation protocols and perspectives
• InteractRank: Personalized Web-Scale Search Pre-Ranking with Cross Interaction Features
• MemoNet: Memorizing All Cross Features' Representations Efficiently via Multi-Hash Codebook Network for CTR Prediction
• Dual-Stream MLP is All You Need for CTR Prediction
• AdSEE: Investigating the Impact of Image Style Editing on Advertisement Attractiveness
• RAMP: Robust Ad Recommendation Under Limited Personalized-Feature Availability via Masking and Alignment Pathways