Development And Application Of SNA-based Virtual Community Group Structure Analysis System

Vigorous development of virtual communities is based on virtual communitycommunication has aggregation is not simply the pairwise communication. Studies have shownthat group members know the group structure will help them to do the conscious knowledgesharing behavior, thereby improving the efficiency of group knowledge sharing. In view of this,this study want to research on members in virtual community knowledge communication, toidentify knowledge-sharing law, to propose a group structure analysis model, then design andimplement a structure analysis system on top of these. The system help the members incommunity to be easy to understand the community structure, and ultimately to improve theefficiency of the virtual community knowledge sharing.First, this paper introduces literature research for the virtual community group structureanalysis system from theoretical basis, existing systems, and core technology. The theoreticalbasis includes virtual community and social network analysis; the existing systems research forthe advantages and disadvantages in UCINET and KIWI; core technology is layout algorithm,mainly analyze force directed layout algorithm and Nbody algorithm.Create a group structure analysis model on the basis of the literature research, to guidedesign and implementation of the group structure system. The model is primarily based on thecommunity knowledge-sharing model between community members and description indexes ofknowledge sharing. the community knowledge-sharing model analyze the virtual communitygroup composition and the communication method in BBS, blogs, wiki. Description indexesinclude members’ participation and roles, group entirety and cohesion.Based on the analysis model this paper gives the system requirements: members’participation and roles, group entirety and cohesion, and multiple layout visualization graphs. Then main function of the virtual community group structure analysis system:(1) calculation ofmeasure indicators, which include: the link strength, density, cohesion sub-group, intermediarydegree, centrality degree;(2) multiple layout visualization graphs, which includes four graphs:directed graph, no-directed graph, subgroup graph and core location view. Then use the WaterfallModel in order to complete the system design and implementation. Then detailed design thesubgroup graph and core location view, which showing small groups and members networklocation these two group structural features.Finally, by two sets of experiments, using the reality teaching data to get operatingconditions and effect of the group structure analysis system. And summarizes the results andproblems, and propose directions for future research.