2022

Breaking Feedback Loops in Recommender Systems with Causal Inference

Krauth, Karl, Wang, Yixin, Jordan, Michael I.

Understand

Recommender systems play a key role in shaping modern web ecosystems.

These systems alternate between (1) making recommendations (2) collecting user responses to these recommendations, and (3) retraining the recommendation algorithm based on this feedback.
During this process the recommender system influences the user behavioral data that is subsequently used to update it, thus creating a feedback loop.
Recent work has shown that feedback loops may compromise recommendation quality and homogenize user behavior, raising ethical and performance concerns when deploying recommender systems.

Built on

The influence of search engines on preferential attachment
Soumen Chakrabarti, Alan Frieze, and Juan Vera · 2006
Earlier work this paper cites.
Experimental study of inequality and unpredictability in an artificial cultural market
Matthew J Salganik, Peter Sheridan Dodds, and Duncan J Watts · 2006
Earlier work this paper cites.
Unifying user-based and item-based collaborative filtering approaches by similarity fusion
Jun Wang, Arjen P De Vries, and Marcel JT Reinders · 2006
Earlier work this paper cites.
Scalable collaborative filtering with jointly derived neighborhood interpolation weights
Robert M Bell and Yehuda Koren · 2007
Earlier work this paper cites.
Collaborative filtering for implicit feedback datasets
Yifan Hu, Yehuda Koren, and Chris Volinsky · 2008
Earlier work this paper cites.
Probabilistic matrix factorization
Andriy Mnih and Ruslan R Salakhutdinov · 2008
Earlier work this paper cites.
Collaborative filtering with temporal dynamics
Yehuda Koren · 2009
Earlier work this paper cites.
Collaborative prediction and ranking with non-random missing data
Benjamin M Marlin and Richard S Zemel · 2009
Earlier work this paper cites.
Causality
Judea Pearl · 2009
Earlier work this paper cites.
Slim: Sparse linear methods for top-n recommender systems
Xia Ning and George Karypis · 2011
Earlier work this paper cites.
Factorization machines with libFM
Steffen Rendle · 2012
Earlier work this paper cites.
Content-based recommendations with poisson factorization
Prem K Gopalan, Laurent Charlin, and David Blei · 2014
Earlier work this paper cites.
Probabilistic matrix factorization with non-random missing data
José Miguel Hernández-Lobato, Neil Houlsby, and Zoubin Ghahramani · 2014
Earlier work this paper cites.
Scalable recommendation with hierarchical poisson factorization
Prem Gopalan, Jake M Hofman, and David M Blei · 2015
Earlier work this paper cites.

Similar

The movielens datasets: History and context
F Maxwell Harper and Joseph A Konstan · 2015
Cited alongside, same era.
Causal inference in statistics, social, and biomedical sciences
Guido W Imbens and Donald B Rubin · 2015
Cited alongside, same era.
Autorec: Autoencoders meet collaborative filtering
Suvash Sedhain, Aditya Krishna Menon, Scott Sanner, and Lexing Xie · 2015
Cited alongside, same era.
Estimating the causal impact of recommendation systems from observational data
Amit Sharma, Jake M Hofman, and Duncan J Watts · 2015
Cited alongside, same era.
LLORMA: Local low-rank matrix approximation
Joonseok Lee, Seungyeon Kim, Guy Lebanon, Yoram Singer, and Samy Bengio · 2016
Cited alongside, same era.
Modeling user exposure in recommendation
Dawen Liang, Laurent Charlin, James McInerney, and David M Blei · 2016
Cited alongside, same era.
Recommendations as treatments: Debiasing learning and evaluation
Tobias Schnabel, Adith Swaminathan, Ashudeep Singh, Navin Chandak, and Thorsten Joachims · 2016
Cited alongside, same era.
Deconvolving feedback loops in recommender systems
Ayan Sinha, David F Gleich, and Karthik Ramani · 2016
Cited alongside, same era.
A neural autoregressive approach to collaborative filtering
Original
Yin Zheng, Bangsheng Tang, Wenkui Ding, and Hanning Zhou · 2016
Cited alongside, same era.
Neural collaborative filtering
Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua · 2017
Cited alongside, same era.
Causal embeddings for recommendation
Stephen Bonner and Flavian Vasile · 2018
Cited alongside, same era.
How algorithmic confounding in recommendation systems increases homogeneity and decreases utility
Allison JB Chaney, Brandon M Stewart, and Barbara E Engelhardt · 2018
Cited alongside, same era.

Then

Human interaction with recommendation systems
Sven Schmit and Carlos Riquelme · 2018
Later among the works it cites.
Degenerate feedback loops in recommender systems
Ray Jiang, Silvia Chiappa, Tor Lattimore, András György, and Pushmeet Kohli · 2019
Later among the works it cites.
Embarrassingly shallow autoencoders for sparse data
Harald Steck · 2019
Later among the works it cites.
Debiasing the human-recommender system feedback loop in collaborative filtering
Wenlong Sun, Sami Khenissi, Olfa Nasraoui, and Patrick Shafto · 2019
Later among the works it cites.
Equal opportunity and affirmative action via counterfactual predictions
Original
Yixin Wang, Dhanya Sridhar, and David M Blei · 2019
Later among the works it cites.
Evaluating the scale, growth, and origins of right-wing echo chambers on youtube
Original
Homa Hosseinmardi, Amir Ghasemian, Aaron Clauset, David M Rothschild, Markus Mobius, and Duncan J Watts · 2020
Later among the works it cites.
Do offline metrics predict online performance in recommender systems?
Original
Karl Krauth, Sarah Dean, Alex Zhao, Wenshuo Guo, Mihaela Curmei, Benjamin Recht, and Michael I Jordan · 2020
Later among the works it cites.
Feedback loop and bias amplification in recommender systems
Masoud Mansoury, Himan Abdollahpouri, Mykola Pechenizkiy, Bamshad Mobasher, and Robin Burke · 2020
Later among the works it cites.
Performative prediction
Juan Perdomo, Tijana Zrnic, Celestine Mendler-Dünner, and Moritz Hardt · 2020
Later among the works it cites.
Causal inference for recommender systems
Yixin Wang, Dawen Liang, Laurent Charlin, and David M Blei · 2020
Later among the works it cites.
On statistical bias in active learning: How and when to fix it
Original
Sebastian Farquhar, Yarin Gal, and Tom Rainforth · 2021
Later among the works it cites.
Preference amplification in recommender systems
Dimitris Kalimeris, Smriti Bhagat, Shankar Kalyanaraman, and Udi Weinsberg · 2021
Later among the works it cites.
Correcting the user feedback-loop bias for recommendation systems
Original
Weishen Pan, Sen Cui, Hongyi Wen, Kun Chen, Changshui Zhang, and Fei Wang · 2021
Later among the works it cites.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…