Abstract
Information stored in online communities consist not only knowledge contents, but also the information of knowledge providers and searchers' connective relationships, and network structures. Online Communities provide effective platforms for interaction and play pivotal roles in making provision for the basis of analysis as all the ask-response paired relationships are automatically recorded. This paper demonstrates how to apply social network analysis to analyze the interaction data for generating the "role information" of the knowledge searchers and providers. Integrating concepts of uncertainty in knowledge searching and sociometric used in social network analysis, we develop a mechanism for role matching in knowledge search for each questions posed. Roles identified in this approach including central, network entrepreneur (e.g. spanning structural holes), neighboring mediate (e.g. knowledge gate keeper), and resource competitor (e.g. structural equivalent players). The result is demonstrated and visualized in a web-based community platform and tested in a real-world programmer forum-based community.