Disentangling User Interest and Conformity for Recommendation with Causal Embedding

Recommendation models are usually trained on observational interaction data. However, observational interaction data could result from users’ conformity towards popular items, which entangles users’ real interest. Existing methods tracks this problem as eliminating popularity bias, e.g., by re-weighting training samples or leveraging a small fraction of unbiased data. However, the variety of user conformity is ignored by these approaches, and different causes of an interaction are bundled together as unified representations, hence robustness and interpretability are not guaranteed when underlying causes are changing. In this paper, we present DICE, a general framework that learns representations where interest and conformity are structurally disentangled, and various backbone recommendation models could be smoothly integrated. We assign users and items with separate embeddings for interest and conformity, and make each embedding capture only one cause by training with cause-specific data which is obtained according to the colliding effect of causal inference. Our proposed methodology outperforms state-of-the-art baselines with remarkable improvements on two real-world datasets on top of various backbone models. We further demonstrate that the learned embeddings successfully capture the desired causes, and show that DICE guarantees the robustness and interpretability of recommendation.

[1]  Barbara E. Engelhardt,et al.  How algorithmic confounding in recommendation systems increases homogeneity and decreases utility , 2017, RecSys.

[2]  James Bennett,et al.  The Netflix Prize , 2007 .

[3]  Christopher Joseph Pal,et al.  A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms , 2019, ICLR.

[4]  Hongxia Yang,et al.  Learning Disentangled Representations for Recommendation , 2019, NeurIPS.

[5]  Lars Schmidt-Thieme,et al.  BPR: Bayesian Personalized Ranking from Implicit Feedback , 2009, UAI.

[6]  Yinghai Lu,et al.  Deep Learning Recommendation Model for Personalization and Recommendation Systems , 2019, ArXiv.

[7]  Jure Leskovec,et al.  Inductive Representation Learning on Large Graphs , 2017, NIPS.

[8]  Maria L. Rizzo,et al.  Measuring and testing dependence by correlation of distances , 2007, 0803.4101.

[9]  Jure Leskovec,et al.  Graph Convolutional Neural Networks for Web-Scale Recommender Systems , 2018, KDD.

[10]  Richard S. Zemel,et al.  Collaborative Filtering and the Missing at Random Assumption , 2007, UAI.

[11]  Harald Steck,et al.  Item popularity and recommendation accuracy , 2011, RecSys '11.

[12]  Alejandro Bellogín,et al.  Statistical biases in Information Retrieval metrics for recommender systems , 2017, Information Retrieval Journal.

[13]  Christopher Burgess,et al.  beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework , 2016, ICLR 2016.

[14]  Mélanie Frappier,et al.  The Book of Why: The New Science of Cause and Effect , 2018, Science.

[15]  David M. Blei,et al.  Causal Inference for Recommendation , 2016 .

[16]  Stefan Bauer,et al.  Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness , 2018, ICML.

[17]  A KonstanJoseph,et al.  The MovieLens Datasets , 2015 .

[18]  Pablo Castells,et al.  A Probabilistic Reformulation of Memory-Based Collaborative Filtering: Implications on Popularity Biases , 2017, SIGIR.

[19]  David M. Blei,et al.  The Deconfounded Recommender: A Causal Inference Approach to Recommendation , 2018, ArXiv.

[20]  Yongdong Zhang,et al.  LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation , 2020, SIGIR.

[21]  Dietmar Jannach,et al.  What recommenders recommend: an analysis of recommendation biases and possible countermeasures , 2015, User Modeling and User-Adapted Interaction.

[22]  Emilien Dupont,et al.  Joint-VAE: Learning Disentangled Joint Continuous and Discrete Representations , 2018, NeurIPS.

[23]  Bamshad Mobasher,et al.  Controlling Popularity Bias in Learning-to-Rank Recommendation , 2017, RecSys.

[24]  Robert M. Bell,et al.  The BellKor 2008 Solution to the Netflix Prize , 2008 .

[25]  Yong Yu,et al.  Are You Influenced by Others When Rating?: Improve Rating Prediction by Conformity Modeling , 2016, RecSys.

[26]  Ben Carterette,et al.  Offline Evaluation to Make Decisions About PlaylistRecommendation Algorithms , 2019, WSDM.

[27]  Yehuda Koren,et al.  Matrix Factorization Techniques for Recommender Systems , 2009, Computer.

[28]  Jason Weston,et al.  Curriculum learning , 2009, ICML '09.

[29]  Bernhard Schölkopf,et al.  Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations , 2018, ICML.

[30]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[31]  Tor Lattimore,et al.  Degenerate Feedback Loops in Recommender Systems , 2019, AIES.

[32]  Bernhard Scholkopf Causality for Machine Learning , 2019 .

[33]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[34]  Pablo Castells,et al.  Should I Follow the Crowd?: A Probabilistic Analysis of the Effectiveness of Popularity in Recommender Systems , 2018, SIGIR.

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

[36]  F. Maxwell Harper,et al.  The MovieLens Datasets: History and Context , 2016, TIIS.

[37]  Nikos Komodakis,et al.  Unsupervised Representation Learning by Predicting Image Rotations , 2018, ICLR.

[38]  Xiangnan He,et al.  Bias and Debias in Recommender System: A Survey and Future Directions , 2020, ACM Trans. Inf. Syst..

[39]  Thorsten Joachims,et al.  A General Framework for Counterfactual Learning-to-Rank , 2018, SIGIR.

[40]  Juan Carlos Niebles,et al.  Learning to Decompose and Disentangle Representations for Video Prediction , 2018, NeurIPS.

[41]  Max Welling,et al.  Auto-Encoding Variational Bayes , 2013, ICLR.

[42]  Punam Bedi,et al.  Using novelty score of unseen items to handle popularity bias in recommender systems , 2014, 2014 International Conference on Contemporary Computing and Informatics (IC3I).

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

[44]  Deborah Estrin,et al.  Unbiased offline recommender evaluation for missing-not-at-random implicit feedback , 2018, RecSys.

[45]  Thomas Nedelec,et al.  Offline A/B Testing for Recommender Systems , 2018, WSDM.

[46]  Stephen Bonner,et al.  Causal embeddings for recommendation , 2017, RecSys.

[47]  Thorsten Joachims,et al.  Recommendations as Treatments: Debiasing Learning and Evaluation , 2016, ICML.

[48]  Joaquin Quiñonero Candela,et al.  Counterfactual reasoning and learning systems: the example of computational advertising , 2013, J. Mach. Learn. Res..