Title: A study of friend recommendations for gaming communities
Authors: Bryan Watson; Thomas Watson; Jun Zheng
Addresses: Department of Computer Science and Engineering, New Mexico Institute of Mining and Technology, Socorro, NM, USA ' Department of Computer Science and Engineering, New Mexico Institute of Mining and Technology, Socorro, NM, USA ' Department of Computer Science and Engineering, New Mexico Institute of Mining and Technology, Socorro, NM, USA
Abstract: Players of online gaming communities such as Steam, Xbox Live, and the PlayStation Network may have trouble finding people to play with as evidenced by the popularity of looking for group (LFG) services. This paper studies friend recommendation systems as a possible solution to alleviate this problem because a high quality friend list can provide a higher chance for the player to find people to play with. An online survey of video game players was conducted to study the need for friend recommendations in gaming communities and how players build their friend lists. The results showed that a sizable portion of players experienced some sort of difficulty finding people to play with and players add friends through a diverse set of possible sources. As the first step to build a friend recommendation system for gaming communities, we tested ten common link prediction similarity indices on a dataset collected from the Xbox Live network containing over 42 million unique users. The results showed that a friend recommendation system-based solely on network topology features did not perform well. Future research should incorporate other information such as player profiles to improve the friend recommendation performance.
Keywords: friend recommendation; gaming communities; looking for group; LFG; link prediction; recommender system.
DOI: 10.1504/IJWBC.2019.103188
International Journal of Web Based Communities, 2019 Vol.15 No.4, pp.292 - 314
Received: 05 Feb 2019
Accepted: 07 Mar 2019
Published online: 21 Oct 2019 *