A fairness aware service recommendation method in service ecosystem

With the rapid development of internet technology, the number of services with the same or similar functions on the internet has increased explosively. How to provide users with more accurate service recommendation is one of the hot issues in academia and industry. However, most of the existing recommendation methods tend to recommend popular services to users, which result into serious polarisation and become a barrier for the unpopular services to startup and growth. To solve this problem, we propose a fairness aware service recommendation (FASR), which pays attention to the fair treatment of unpopular services in the process of service recommendation. FASR addresses both accuracy and fairness, and designs different recommendation algorithms for popular and unpopular services respectively. A large number of experiments and analyses show that FASR can significantly improve the fairness of recommendations with little impact on accuracy in the evolving service ecosystem.