Kalman Filtering Attention for User Behavior Modeling in CTR Prediction

Click-through rate (CTR) prediction is one of the fundamental tasks for e-commerce search engines. As search becomes more personalized, it is necessary to capture the user interest from rich behavior data. Existing user behavior modeling algorithms develop different attention mechanisms to emphasize query-relevant behaviors and suppress irrelevant ones. Despite being extensively studied, these attentions still suffer from two limitations. First, conventional attentions mostly limit the attention field only to a single user's behaviors, which is not suitable in e-commerce where users often hunt for new demands that are irrelevant to any historical behaviors. Second, these attentions are usually biased towards frequent behaviors, which is unreasonable since high frequency does not necessarily indicate great importance. To tackle the two limitations, we propose a novel attention mechanism, termed Kalman Filtering Attention (KFAtt), that considers the weighted pooling in attention as a maximum a posteriori (MAP) estimation. By incorporating a priori, KFAtt resorts to global statistics when few user behaviors are relevant. Moreover, a frequency capping mechanism is incorporated to correct the bias towards frequent behaviors. Offline experiments on both benchmark and a 10 billion scale real production dataset, together with an Online A/B test, show that KFAtt outperforms all compared state-of-the-arts. KFAtt has been deployed in the ranking system of a leading e commerce website, serving the main traffic of hundreds of millions of active users everyday.

[1]  Eunho Yang,et al.  Uncertainty-Aware Attention for Reliable Interpretation and Prediction , 2018, NeurIPS.

[2]  T. Başar,et al.  A New Approach to Linear Filtering and Prediction Problems , 2001 .

[3]  Hongtao Lu,et al.  Deep CTR Prediction in Display Advertising , 2016, ACM Multimedia.

[4]  Li Zhang,et al.  Field-Aware Neural Factorization Machine for Click-Through Rate Prediction , 2019, IEEE Access.

[5]  Paul Covington,et al.  Deep Neural Networks for YouTube Recommendations , 2016, RecSys.

[6]  Anton van den Hengel,et al.  Image-Based Recommendations on Styles and Substitutes , 2015, SIGIR.

[7]  Chang Zhou,et al.  Deep Interest Evolution Network for Click-Through Rate Prediction , 2018, AAAI.

[8]  Guorui Zhou,et al.  Practice on Long Sequential User Behavior Modeling for Click-Through Rate Prediction , 2019, KDD.

[9]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[10]  Feng Yu,et al.  A Dynamic Recurrent Model for Next Basket Recommendation , 2016, SIGIR.

[11]  T. Bayes.  An essay towards solving a problem in the doctrine of chances , 2003 .

[12]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[13]  Joaquin Quiñonero Candela,et al.  Practical Lessons from Predicting Clicks on Ads at Facebook , 2014, ADKDD'14.

[14]  Martin Wattenberg,et al.  Ad click prediction: a view from the trenches , 2013, KDD.

[15]  Xueqi Cheng,et al.  Learning to Control the Specificity in Neural Response Generation , 2018, ACL.

[16]  Yoshua Bengio,et al.  Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling , 2014, ArXiv.

[17]  Guorui Zhou,et al.  Deep Interest Network for Click-Through Rate Prediction , 2017, KDD.

[18]  Qiang Yang,et al.  Personalized click model through collaborative filtering , 2012, WSDM '12.

[19]  Heng-Tze Cheng,et al.  Wide & Deep Learning for Recommender Systems , 2016, DLRS@RecSys.

[20]  Yang Song,et al.  Multi-Rate Deep Learning for Temporal Recommendation , 2016, SIGIR.

[21]  Chang Zhou,et al.  ATRank: An Attention-Based User Behavior Modeling Framework for Recommendation , 2017, AAAI.

[22]  Yoshua Bengio,et al.  Neural Machine Translation by Jointly Learning to Align and Translate , 2014, ICLR.

[23]  Philip K. Chan,et al.  Learning implicit user interest hierarchy for context in personalization , 2008, IUI '03.

[24]  Keping Yang,et al.  Deep Session Interest Network for Click-Through Rate Prediction , 2019, IJCAI.

[25]  Xiaodong He,et al.  A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems , 2015, WWW.

[26]  Yann Dauphin,et al.  Convolutional Sequence to Sequence Learning , 2017, ICML.

[27]  Yu Zhang,et al.  Image Matters: Visually Modeling User Behaviors Using Advanced Model Server , 2017, CIKM.

[28]  Torsten Suel,et al.  Modeling and predicting user behavior in sponsored search , 2009, KDD.

[29]  Yunming Ye,et al.  DeepFM: A Factorization-Machine based Neural Network for CTR Prediction , 2017, IJCAI.

[30]  Matthew Richardson,et al.  Predicting clicks: estimating the click-through rate for new ads , 2007, WWW '07.