Collaborative Filtering Prone to Popularity Bias, Resulting in Overrepresentation of Popular Items in the Recommendation Outputs

March 1, 2022

Collaborative filtering is often employed in recommender systems, but it can be prone to popularity bias. This occurs when popular items receive more recommendations than less popular ones, leading to overrepresentation of these items in the output. As a result, users may find themselves being recommended similar products or content that they have already seen before, rather than discovering new and relevant options.

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

Alleged deployer
YouTube, Twitter, Netflix, LinkedIn, Facebook
Alleged developer
YouTube, Twitter, Netflix, LinkedIn, Facebook
Alleged harmed parties
YouTube users, X (Twitter) users, Netflix users, LinkedIn users, Facebook users

AI governance case studies

For forensic AI governance failure analysis (TAIMScore™ case studies), browse Human Signal’s Failure Files™.

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/168

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.