Organization Based On Complex Network Analysis

In the age of explosive growth of data, analysis and mining of information has become one of today’s organization of management and information science study on the hot and difficult. Nature of the data is from an existing data mining of valuable information and analysis, so how to apply theory related to large data to related disciplines is a process that requires constant exploration. But most of the existing research is based on a social network relationship forecast, forecast public opinion monitoring or social events, the lack of analysis of the relationship between departments in the Organization of research. Based on the complex network analysis methods and ideas, and Organization study on the analytical method for the design and tries to unlock the organization cooperation relationship between the internal departments and features.Study on analytical methods for organization, because the Organization modes are similar, and sensitivity to other organizations, therefore this article targeted to organizations in colleges and universities. This study on the organization method based on complex networks, major contributions is as follows:1. Based on the complex network of outside news organizations build method. This method is used primarily by the actual organizational structure(University) crawling the news site, gets its related news, news analysis and clean up the data, eliminating irrelevant news data, then using entity relation extraction method for extracting relationships between departments in the news page. Then the relationships between all the sectors are combined, to build complex organization network.2. Presents label fusion communities Division of analytical method of organization.First, the Organization of complex network application associations label Division of fusion algorithm to be communities, through times divided, gets to organization institutions internal connection most close of community, through complex network in the of degrees and side weight of analysis method on its for analytical, and compared two a similar Yes organization institutions between of analytical results, to validation results of reliability,eventually proved similar of organization institutions of internal of cooperation mode is similar like.3. Node merging method based on seed faction is proposed to resolve organizational network. Find complex networks in a fully connected network, its factions as seeds, then using hierarchical clustering methods, layer-by-layer construction of the organization network, and UESTC organizations network with Xidian University organization network of similarities and differences.In summary, data through news UESTC organizations network with Xidian University organization complex networks, and its analytical comparison, verified by analysis of complex network analysis methods to the organization network, to gain access to hidden relations within the Agency.