Social Network Analysis on Communications for Knowledge Collaboration in OSS Communities

Knowledge collaboration is the key for success of open source software (OSS) communities, because not all members have knowledge and skills necessary for software development. Generally, members in OSS communities communicate for knowledge collaboration using communication tools (e.g. mailing lists, discussion forums, bug tracking systems, and so on) so that geographically distributed members collaborate and coordinate their work. In this paper, we apply social network analysis to the data accumulated in communication tools. We analyzed relationships between the density of social networks and OSS releases by time series analysis of 4 OSS communities in SourceForge.net in order to investigate the quality of communications for knowledge collaboration. The analysis results showed that communications among community members with a variety of roles are active before/after OSS released in communities where knowledge collaboration is going well.