Recommending what video to watch next: a multitask ranking system

In this paper, we introduce a large scale multi-objective ranking system for recommending what video to watch next on an industrial video sharing platform. The system faces many real-world challenges, including the presence of multiple competing ranking objectives, as well as implicit selection biases in user feedback. To tackle these challenges, we explored a variety of soft-parameter sharing techniques such as Multi-gate Mixture-of-Experts so as to efficiently optimize for multiple ranking objectives. Additionally, we mitigated the selection biases by adopting a Wide & Deep framework. We demonstrated that our proposed techniques can lead to substantial improvements on recommendation quality on one of the world's largest video sharing platforms.

[1]  Xin Zhang,et al.  TFX: A TensorFlow-Based Production-Scale Machine Learning Platform , 2017, KDD.

[2]  Ed H. Chi,et al.  Towards Neural Mixture Recommender for Long Range Dependent User Sequences , 2019, WWW.

[3]  Yue Yin,et al.  Explainable Recommendation via Multi-Task Learning in Opinionated Text Data , 2018, SIGIR.

[4]  Zhe Zhao,et al.  Improving User Topic Interest Profiles by Behavior Factorization , 2015, WWW.

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

[6]  Boi Faltings,et al.  Offline and online evaluation of news recommender systems at swissinfo.ch , 2014, RecSys '14.

[7]  Thorsten Joachims,et al.  Unbiased Learning-to-Rank with Biased Feedback , 2016, WSDM.

[8]  Jimmy J. Lin,et al.  WTF: the who to follow service at Twitter , 2013, WWW.

[9]  Gang Fu,et al.  Deep & Cross Network for Ad Click Predictions , 2017, ADKDD@KDD.

[10]  Filip Radlinski,et al.  Evaluating the accuracy of implicit feedback from clicks and query reformulations in Web search , 2007, TOIS.

[11]  Mohamed A. Sharaf,et al.  MuVE: Efficient Multi-Objective View Recommendation for Visual Data Exploration , 2016, 2016 IEEE 32nd International Conference on Data Engineering (ICDE).

[12]  George Karypis,et al.  Multi-task learning for recommender systems , 2010, ACML 2010.

[13]  Ji-Rong Wen,et al.  A Neural Labeled Network Embedding Approach to Product Adopter Prediction , 2018, AIRS.

[14]  Philip S. Yu,et al.  Learning Multiple Tasks with Multilinear Relationship Networks , 2015, NIPS.

[15]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[16]  Zhe Zhao,et al.  Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations , 2017, ArXiv.

[17]  Barry Smyth,et al.  Why I like it: multi-task learning for recommendation and explanation , 2018, RecSys.

[18]  Ed H. Chi,et al.  SNR: Sub-Network Routing for Flexible Parameter Sharing in Multi-Task Learning , 2019, AAAI.

[19]  Tie-Yan Liu,et al.  LightGBM: A Highly Efficient Gradient Boosting Decision Tree , 2017, NIPS.

[20]  Trevor Darrell,et al.  Visual Discovery at Pinterest , 2017, WWW.

[21]  Trevor Darrell,et al.  Adversarial Discriminative Domain Adaptation , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Geoffrey E. Hinton,et al.  Adaptive Mixtures of Local Experts , 1991, Neural Computation.

[23]  Zhe Zhao,et al.  Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts , 2018, KDD.

[24]  Marc Najork,et al.  Learning to Rank with Selection Bias in Personal Search , 2016, SIGIR.

[25]  Geoffrey E. Hinton,et al.  Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer , 2017, ICLR.

[26]  Olivier Chapelle,et al.  A dynamic bayesian network click model for web search ranking , 2009, WWW '09.

[27]  Stephanie Rogers,et al.  Related Pins at Pinterest: The Evolution of a Real-World Recommender System , 2017, WWW.

[28]  Maoguo Gong,et al.  Multi-objective optimization for long tail recommendation , 2016, Knowl. Based Syst..

[29]  George Karypis,et al.  Multi-task Learning for Recommender System , 2010, ACML.

[30]  Thorsten Joachims,et al.  Estimating Position Bias without Intrusive Interventions , 2018, WSDM.

[31]  Lina Yao,et al.  Deep Learning Based Recommender System , 2017, ACM Comput. Surv..

[32]  Thorsten Joachims,et al.  Batch learning from logged bandit feedback through counterfactual risk minimization , 2015, J. Mach. Learn. Res..

[33]  Jianmin Wang,et al.  Learning Multiple Tasks with Deep Relationship Networks , 2015, ArXiv.

[34]  Jure Leskovec,et al.  Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time , 2017, WWW.

[35]  Deepak Agarwal,et al.  Localized factor models for multi-context recommendation , 2011, KDD.

[36]  Yu He,et al.  The YouTube video recommendation system , 2010, RecSys '10.

[37]  Karthik Ramani,et al.  Deconvolving Feedback Loops in Recommender Systems , 2016, NIPS.

[38]  Antonino Freno,et al.  Practical Lessons from Developing a Large-Scale Recommender System at Zalando , 2017, RecSys.

[39]  Zhoujun Li,et al.  Integrating rich information for video recommendation with multi-task rank aggregation , 2011, ACM Multimedia.

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

[41]  John R. Anderson,et al.  Efficient Training on Very Large Corpora via Gramian Estimation , 2018, ICLR.

[42]  Abien Fred Agarap.  Deep Learning using Rectified Linear Units (ReLU) , 2018, ArXiv.

[43]  Ke Wang,et al.  Ranking Distillation: Learning Compact Ranking Models With High Performance for Recommender System , 2018, KDD.

[44]  Gregory N. Hullender,et al.  Learning to rank using gradient descent , 2005, ICML.

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

[46]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

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