Research On Transactional Community's Structure And Evolution Based On Social Network

With the rapid development and popularity of Internet technology, the transactional network community gradually shows strong marketing function and potential business value. Research on consumers'behavior of network community can provide important basis for decision-making enterprise. At present social network analysis of the online community is mainly on nonprofit community, such as music BBS community, BLOG community. Research on transactional community aimed at profit is less. Taking this into consideration, the paper choosed Taobao website, the China's largest C2C platform website, for study. We put the social network analysis theory into the research on "adorn article/popular jewelry/fashion jewelry". The research was did from the static analysis of individual's structure, the characteristic analysis of the overall structure and the dynamic evolution analysis of the whole structure three aspects deeply and systematically.First of all, the paper made the static analysis of individual's structure from degree centrality, betweenness centrality and closeness centrality three aspects. The study found that there was a certain degree of group centralization phenomenon in the network. The vast majority of nodes had no control of resources and control only concentrated in a few hands. Moreover, the reliance between nodes was strong relatively. Second, the properties of the whole network were analyzed from close degree of relation between nodes, small world characteristic, scale-free characteristic, clique property and with match property. The result indicated that contact between nodes was loose, small world property insignificant and the clique phenomenon not serious. At the same time, the network had the property of scale-free and different match. At last, in the section of network dynamic evolution, network models of ten stages based on core layer elements which were mined by use of the improved algorithm of CNKM were constructed respectively. And then, we described the dynamic changeable process of actual network from degree distribution, clustering coefficient and average path length separately. Based on the description of actual network, the paper proposed an evolution model based on relations between existing nodes and verified the validity of the evolution model in the way of computer simulation.The research, on the one hand, enriched the contents of social network analysis theory. On the other hand, it is helpful in fully grasping network characteristics of communities of this type and flow direction of network resources. At the same time, it can provide important references and basis for network operators'making more conductive development decision, coordinately, promoting the development of C2C e-business.