FiBiNET: combining feature importance and bilinear feature interaction for click-through rate prediction

@article{Huang2019FiBiNETCF,
  title={FiBiNET: combining feature importance and bilinear feature interaction for click-through rate prediction},
  author={Tongwen Huang and Zhiqi Zhang and Junlin Zhang},
  journal={Proceedings of the 13th ACM Conference on Recommender Systems},
  year={2019}
}
Advertising and feed ranking are essential to many Internet companies such as Facebook and Sina Weibo. [] Key Method 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…

Figures and Tables from this paper

Leaf-FM: A Learnable Feature Generation Factorization Machine for Click-Through Rate Prediction

This paper proposes LeafFM model based on FM to generate new features from the original feature embedding by learning the transformation functions automatically, and designs three concrete Leaf-FM models according to the different strategies of combing the original and the generated features.

GateNet: Gating-Enhanced Deep Network for Click-Through Rate Prediction

A novel model named GateNet is proposed which introduces either the feature embedding gate or the hidden gate to the embedding layer or hidden layers of DNN CTR models, respectively to boost the performance of various state-of-the-art models such as FM, DeepFM and xDeepFM on all datasets.

Feature Interaction based Neural Network for Click-Through Rate Prediction

A Feature Interaction based Neural Network (FINN) which is able to model feature interaction via a 3-dimention relation tensor and provides representations for the feature interactions on the bottom layer and the non-linearity of neural network in modelling higher-order feature interactions.

An Efficient Deep Interaction Network for Click-Through Rate Prediction

A deep interaction network is proposed in this paper, in which the 1-order, 2-Order, 3- order, 4-order features interaction behavior are explicitly modelled and a convolutional based feature extractor for learning high order features combination is integrated in parallel.

Hierarchical attention and feature projection for click-through rate prediction

A framework called Hierarchical Attention and Feature Projection neural network (HAFP) is proposed, which enables the automatically learning of more representative and efficient feature representation in an end-to-end manner and outperforms the state-of-the-art in terms of Logloss and AUC for CTR prediction baselines.

GAIN: A Gated Adaptive Feature Interaction Network for Click-Through Rate Prediction

This work proposes a lightweight, yet effective model called the Gated Adaptive feature Interaction Network (GAIN), and devise a novel cross module to drop meaningless feature interactions and preserve informative ones and transfers learned interactions to other models, such as Logistic Regression and Factorization Machines, and finds out that their performance can be significantly improved.

Exploring different interaction among features for CTR prediction

This paper proposes a CTR prediction model based on multi-feature interaction, called EDIF, which aims to enhance the diversity of feature interaction and shows that the model effect of this paper has great advantages over the latest model.

Se-xDeepFEFM: Combining Low-Order Feature Refinement and Interaction Intensity Evaluation for Click-Through Rate Prediction

A novel model called Senet and extreme deep field-embedded factorization machine (Se-xDFEFM) is proposed for more effective CTR prediction and achieves competitive recommendation performance, demonstrating their scalability.

JointCTR: a joint CTR prediction framework combining feature interaction and sequential behavior learning

A modular click-through rate prediction framework JointCTR is proposed, which yields competitive performance compared to state-of-the-art models and almost all different types of popular models can be assembled into it, that shows the flexibility and scalability of the framework.

JointCTR: a joint CTR prediction framework combining feature interaction and sequential behavior learning

A modular click-through rate prediction framework JointCTR is proposed, which yields competitive performance compared to state-of-the-art models and almost all different types of popular models can be assembled into it, that shows the flexibility and scalability of the framework.
...

References

SHOWING 1-10 OF 32 REFERENCES

xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems

A novel Compressed Interaction Network (CIN), which aims to generate feature interactions in an explicit fashion and at the vector-wise level and is named eXtreme Deep Factorization Machine (xDeepFM), which is able to learn certain bounded-degree feature interactions explicitly and can learn arbitrary low- and high-order feature interactions implicitly.

Deep Interest Network for Click-Through Rate Prediction

A novel model: Deep Interest Network (DIN) is proposed which tackles this challenge by designing a local activation unit to adaptively learn the representation of user interests from historical behaviors with respect to a certain ad.

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

This paper shows that it is possible to derive an end-to-end learning model that emphasizes both low- and high-order feature interactions, and combines the power of factorization machines for recommendation and deep learning for feature learning in a new neural network architecture.

Deep Learning over Multi-field Categorical Data - - A Case Study on User Response Prediction

This paper proposes two novel models using deep neural networks (DNNs) to automatically learn effective patterns from categorical feature interactions and make predictions of users' ad clicks and demonstrates that their methods work better than major state-of-the-art models.

Field-aware Factorization Machines for CTR Prediction

This paper establishes FFMs as an effective method for classifying large sparse data including those from CTR prediction, and proposes efficient implementations for training FFMs and comprehensively analyze FFMs.

Neural Factorization Machines for Sparse Predictive Analytics

NFM seamlessly combines the linearity of FM in modelling second- order feature interactions and the non-linearity of neural network in modelling higher-order feature interactions, and is more expressive than FM since FM can be seen as a special case of NFM without hidden layers.

Deep & Cross Network for Ad Click Predictions

This paper proposes the Deep & Cross Network (DCN), which keeps the benefits of a DNN model, and beyond that, it introduces a novel cross network that is more efficient in learning certain bounded-degree feature interactions.

Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks

A novel model named Attentional Factorization Machine (AFM), which learns the importance of each feature interaction from data via a neural attention network, which consistently outperforms the state-of-the-art deep learning methods Wide&Deep and DeepCross with a much simpler structure and fewer model parameters.

Wide & Deep Learning for Recommender Systems

Wide & Deep learning is presented---jointly trained wide linear models and deep neural networks---to combine the benefits of memorization and generalization for recommender systems and is open-sourced in TensorFlow.

Practical Lessons from Predicting Clicks on Ads at Facebook

This paper introduces a model which combines decision trees with logistic regression, outperforming either of these methods on its own by over 3%, an improvement with significant impact to the overall system performance.