Abstract
The Internet is a platform embedded with new interaction opportunities that represents a new facet of our society. Social scientists have recognized the richness concealed in the interactions of virtual communities and made endeavors to understand its relation with the physical world. More efforts are devoted to extracting the business value of virtual communities and finding viable business models. However, due to the lack of suitable software, analysis of virtual communities cannot be easily completed. The purpose of this research is to propose an automated tool that can provide analysis for understanding in virtual communities. Specifically, we utilize social network analysis as a theoretical base in developing the system. Social network analysis goes beyond individual-based analysis to describe networks of relations and their impacts. It consists of a framework and network indexes for analyzing relationships in communities. This approach has been used widely by social scientists to investigate the dynamics of social groups. In addition to incorporating the theory, the proposed system features the capabilities of visual analysis, time-dependent analysis, and network structure analysis. This approach can identify and analyze subgroups and elevate the abstraction level of analysis, effectively reducing information overload. By incorporating the network indexes, it can also help identify members' pivotal roles and allow comparisons among networks. The resulting system can be a reference framework for developing a community support platform and help the management of virtual community.