Community Analysis of Social Virtual Reality Based on Large-Scale Log Data of a Commercial Metaverse Platform (2509.23654v1)

Abstract: This study quantitatively analyzes the structural characteristics of user communities within Social Virtual Reality (Social VR) platforms supporting head-mounted displays (HMDs), based on large-scale log data. By detecting and evaluating community structures from data on substantial interactions (defined as prolonged co-presence in the same virtual space), we found that Social VR platforms tend to host numerous, relatively small communities characterized by strong internal cohesion and limited inter-community connections. This finding contrasts with the large-scale, broadly connected community structures typically observed in conventional Social Networking Services (SNS). Furthermore, we identified a user segment capable of mediating between communities, despite these users not necessarily having numerous direct connections. We term this user segment `community hoppers' and discuss their characteristics. These findings contribute to a deeper understanding of the community structures that emerge within the unique communication environment of Social VR and the roles users play within them.