Existence conditions for hidden feedback loops in online recommender systems

09/11/2021
by   Anton S. Khritankov, et al.
0

We explore a hidden feedback loops effect in online recommender systems. Feedback loops result in degradation of online multi-armed bandit (MAB) recommendations to a small subset and loss of coverage and novelty. We study how uncertainty and noise in user interests influence the existence of feedback loops. First, we show that an unbiased additive random noise in user interests does not prevent a feedback loop. Second, we demonstrate that a non-zero probability of resetting user interests is sufficient to limit the feedback loop and estimate the size of the effect. Our experiments confirm the theoretical findings in a simulated environment for four bandit algorithms.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/21/2020

Bandits Under The Influence (Extended Version)

Recommender systems should adapt to user interests as the latter evolve....
research
07/01/2021

The Use of Bandit Algorithms in Intelligent Interactive Recommender Systems

In today's business marketplace, many high-tech Internet enterprises con...
research
02/27/2019

Degenerate Feedback Loops in Recommender Systems

Machine learning is used extensively in recommender systems deployed in ...
research
09/06/2018

Deep neural network marketplace recommenders in online experiments

Recommendations are broadly used in marketplaces to match users with ite...
research
08/16/2019

Accelerated learning from recommender systems using multi-armed bandit

Recommendation systems are a vital component of many online marketplaces...
research
08/21/2020

Theoretical Modeling of the Iterative Properties of User Discovery in a Collaborative Filtering Recommender System

The closed feedback loop in recommender systems is a common setting that...
research
07/26/2019

On the Value of Bandit Feedback for Offline Recommender System Evaluation

In academic literature, recommender systems are often evaluated on the t...

Please sign up or login with your details

Forgot password? Click here to reset