※一覧ページでは各セミナーをクリックまたはタップすることで詳細をご覧いただけます。On the list page, please click or tap on each seminar to find out more details.
- This event has passed.
Yoji Tomita (CyberAgent) Microeconomic Theory Workshop
June 10 @ 10:25 AM - 12:10 PM
Title: Balancing Fairness and High Match Rates in Reciprocal Recommender Systems: A Nash Social Welfare Approach
Abstract:
Matching platforms, such as online dating services that help users find partners and job recommendations that connect job seekers with employers, have become increasingly prevalent. For the success of these matching platforms, it is crucial to design reciprocal recommender systems that not only increase the total number of matches but also avoid creating unfairness among users.
In this paper, we investigate the fairness of reciprocal recommender systems on matching platforms. From the perspective of fair division, we define the users’ opportunities to be recommended and establish the fairness concept of envy-freeness in the allocation of these opportunities. First, we introduce the Social Welfare (SW) method, which approximately maximizes the number of matches, and show that it leads to significant unfairness in recommendation opportunities, illustrating the trade-off between fairness and match rates. Next, we propose the NSW method, which alternately maximizes two Nash social welfare functions and demonstrate that it achieves a fair recommendation with almost zero envy. We further generalize the SW and NSW method to the α-SW method, which balances the trade-off between fairness and high match rates. Furthermore, we develop a fast approximation algorithm for the SW/NSW/α-SW methods based on the Sinkhorn algorithm. Finally, we validate the effectiveness of our approach through three experiments on synthetic data and one experiment on a real-world dataset from a Japanese online dating service.
In this paper, we investigate the fairness of reciprocal recommender systems on matching platforms. From the perspective of fair division, we define the users’ opportunities to be recommended and establish the fairness concept of envy-freeness in the allocation of these opportunities. First, we introduce the Social Welfare (SW) method, which approximately maximizes the number of matches, and show that it leads to significant unfairness in recommendation opportunities, illustrating the trade-off between fairness and match rates. Next, we propose the NSW method, which alternately maximizes two Nash social welfare functions and demonstrate that it achieves a fair recommendation with almost zero envy. We further generalize the SW and NSW method to the α-SW method, which balances the trade-off between fairness and high match rates. Furthermore, we develop a fast approximation algorithm for the SW/NSW/α-SW methods based on the Sinkhorn algorithm. Finally, we validate the effectiveness of our approach through three experiments on synthetic data and one experiment on a real-world dataset from a Japanese online dating service.