Recommendation Systems: Fairness Matters
Abstract (300 words maximum)
There is an increasing focus on fairness in recommender systems, with a growing body of literature on ways to promote fairness. However, this research is fragmented and lacks organization, making it difficult for new researchers to enter the field. Therefore, this survey aims to fill this gap by conducting a thorough analysis of the literature on recommendation fairness. This study focuses on the theoretical underpinnings of fairness in the literature on recommendations. In order to give a general overview of fairness research and to introduce the more complex situations and challenges that must be taken into account when studying fairness in recommender systems, it first presents a brief introduction about fairness in fundamental machine learning tasks such as classification and ranking. The survey will next address fairness in recommendations with an emphasis on taxonomy of current fairness definitions, common methods for enhancing fairness, and datasets for fairness research in recommendations. The study also discusses opportunities and problems in fairness research in an effort to further this field of study along with others.
Academic department under which the project should be listed
CCSE - Computer Science
Primary Investigator (PI) Name
Xinyue Zhang
Recommendation Systems: Fairness Matters
There is an increasing focus on fairness in recommender systems, with a growing body of literature on ways to promote fairness. However, this research is fragmented and lacks organization, making it difficult for new researchers to enter the field. Therefore, this survey aims to fill this gap by conducting a thorough analysis of the literature on recommendation fairness. This study focuses on the theoretical underpinnings of fairness in the literature on recommendations. In order to give a general overview of fairness research and to introduce the more complex situations and challenges that must be taken into account when studying fairness in recommender systems, it first presents a brief introduction about fairness in fundamental machine learning tasks such as classification and ranking. The survey will next address fairness in recommendations with an emphasis on taxonomy of current fairness definitions, common methods for enhancing fairness, and datasets for fairness research in recommendations. The study also discusses opportunities and problems in fairness research in an effort to further this field of study along with others.