Mitigating Mainstream Bias in Recommendation via Cost-sensitive Learning

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

32 Downloads (Pure)

Abstract

Mainstream bias, where some users receive poor recommendations because their preferences are uncommon or simply because they are less active, is an important aspect to consider regarding fairness in recommender systems. Existing methods to mitigate mainstream bias do not explicitly model the importance of these non-mainstream users or, when they do, it is in a way that is not necessarily compatible with the data and recommendation model at hand. In contrast, we use the recommendation utility as a more generic and implicit proxy to quantify mainstreamness, and propose a simple user-weighting approach to incorporate it into the training process while taking the cost of potential recommendation errors into account. We provide extensive experimental results showing that quantifying mainstreamness via utility is better able at identifying non-mainstream users, and that they are indeed better served when training the model in a cost-sensitive way. This is achieved with negligible or no loss in overall recommendation accuracy, meaning that the models learn a better balance across users. In addition, we show that research of this kind, which evaluates recommendation quality at the individual user level, may not be reliable if not using enough interactions when assessing model performance.

Original languageEnglish
Title of host publicationICTIR 2023 - Proceedings of the 2023 ACM SIGIR International Conference on the Theory of Information Retrieval
PublisherAssociation for Computing Machinery (ACM)
Pages135-142
Number of pages8
ISBN (Electronic)9798400700736
DOIs
Publication statusPublished - 2023
Event9th ACM SIGIR International Conference on the Theory of Information Retrieval, ICTIR 2023 - Taipei, Taiwan
Duration: 23 Jul 2023 → …

Publication series

NameICTIR 2023 - Proceedings of the 2023 ACM SIGIR International Conference on the Theory of Information Retrieval

Conference

Conference9th ACM SIGIR International Conference on the Theory of Information Retrieval, ICTIR 2023
Country/TerritoryTaiwan
CityTaipei
Period23/07/23 → …

Keywords

  • bias mitigation
  • mainstream bias
  • recommender systems

Fingerprint

Dive into the research topics of 'Mitigating Mainstream Bias in Recommendation via Cost-sensitive Learning'. Together they form a unique fingerprint.

Cite this