Degenerate Feedback Loops in Recommender Systems

By Ray Jiang, Silvia Chiappa, Tor Lattimore, András György, Pushmeet Kohli. 2019

247 citations recorded by the source corpus.

Machine learning is used extensively in recommender systems deployed in products. The decisions made by these systems can influence user beliefs and preferences which in turn affect the feedback the learning system receives - thus creating a feedback loop. This phenomenon can give rise to the so-called "echo chambers" or "filter bubbles" that have user and societal implications. In this paper, we provide a novel theoretical analysis that examines both the role of user dynamics and the behavior of recommender systems, disentangling the echo chamber from the filter bubble effect. In addition, we offer practical solutions to slow down system degeneracy. Our study contributes toward understanding and developing solutions to commonly cited issues in the complex temporal scenario, an area that is still largely unexplored.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering, Management Science and Operations Research, Evaluation & Benchmarks, Recommender Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Electrical and Electronic Engineering

In Management Science and Operations Research

In Evaluation & Benchmarks

In Recommender Systems

Citations in the corpus

Harms from Increasingly Agentic Algorithmic Systems (2023, background): “…and Gebru 2018; Ehsan et al. 2022; Kasy and Abebe 2021; steven t. piantadosi [@spiantado] 2022; Wolfe and Caliskan 2022], the generation of toxic language [Abid et al. 2021; Gehman et al. 2020], and informational harms [Carroll et al. 2022; Jiang et al. 2019; Li et al. 2022; Weidinger et al. 2022].”
Feedback Loop and Bias Amplification in Recommender Systems (2020, background): “recommender systems has recently received more attention from researchers [4, 5, 9, 14–16].”
Towards Intersectionality in Machine Learning: Including More Identities, Handling Underrepresentation, and Performing Evaluation (2022, background): “…for problems with intersectional concerns may be more akin to that of simulations as testbeds and tools for understanding sociotechical systems such as in modeling recommender systems [17, 21, 45, 70, 89] and online information diffusion [44, 48, 119], rather than as a source for more data.”
Bias and Debias in Recommender System: A Survey and Future Directions (2020, background): “That is, for each user, they do not use the recommendation model for item delivery, but instead randomly select some items and rank them with a uniform distribution [29], [60].”
Model-Agnostic Counterfactual Reasoning for Eliminating Popularity Bias in Recommender System (2020, background): “The main purpose of introducing causal inference in recommender systems is to remove the bias [4, 5, 8, 25, 42, 48, 57].”
Tackling threats to informed decision-making in democratic societies: Promoting epistemic security in a technologically-advanced world (2020, background): “Here, the system may learn the bias of the user and thus create a reinforcing feedback loop whereby users become increasingly exposed to information that coheres with their biases (Jiang et al. 2019).”
Data Feedback Loops: Model-driven Amplification of Dataset Biases (2022, background): “In this context, existing work has shown that optimizing strictly for ranking metrics such as accuracy can create echo chambers, where minority populations are crowded out and disengage from the platform (Hashimoto et al., 2018; Jiang et al., 2019).”