Research On Ontology Based Information Visualization Of Social Network

Information visualization of social network is an important branch of information visualization. It provides intuitionistic and visual displaying of social network information and is an important assistant facility of social network analysis. Information visualization of social network becomes an active research point in the area of information visualization along with the increase of its applications in social network information analysis. Though the information visualization technology of social network has developed rapidly, it still has some problems as the visual algorithm lacking domain knowledge of social network analysis, applications of ontology lacking uniform domain model, the display results of the visual layout algorithm are poor, dense and unreadable, and the bugs of the compress layout algorithms about compress effort isn't so good, correspondences is poor and may lost useful information.Focusing on the problems and needs of information visualization of social network, this paper researchs and analysis the conceptions, methods and tasks in domain of social network analysis and visualization deeply through Ontology method, and proposes a domain ontology model of social network analysis and visualization. Undering this model, we study the analysis methods of social network information, layout display algorithms and compress layout algorithms. Besides, a prototype system of information visualization of social network based on ontology is designed and developed in this paper. The main contributions and work of this paper are listed as following:1. We propose a domain ontology model of social network analysis and visualization - DOSNAV model. Comparing with the ontology models in the area of information visualization of social network, DOSNAV model can summarize and abstract the problems and work of social network analysis and visualization thus shares and reuses the knowledge inside the domain. Accordingly, the DOSNAV model resolves the problems about lacking domain knowledge of social network analysis and lacking uniform domain model.2. We propose a subgroup method of social network analysis - SA method. Undering the DOSNAV model, SA method modifies the methods of social network analysis focusing on the needs of layout and compress layout algorithms. The SA method processing role analysis inside subgroups, subgroup analysis based on regular equivalence groups, subgroup analysis based on key attributes and relative centrality analysis inside subgroup respectively. The analysis results of the SA method are more abundant than those of traditional methods of social network analysis. Thereby they not only lay a foundation for the SAL algorithm and the SALC algorithm, but also facilitate user's better using and understanding of social network information.3. We propose a subgroup analysis layout algorithm - the SAL algorithm for effectively 2D and 3D visualizing of social network information. The SAL algorithm modifies the traditional force directed layout algorithm based on analysis results of SA method. Compareing with traditional force directed layout algorithm, the SAL algorithm can display more abundant results of social network analysis, there by resolves the problem of the information visualization algorithm about poor display results.4. We propose a subgroup analysis layout compress algorithm - the SALC algorithm. The SALC algorithm reduces the result pictures through three levels as regular equivalence groups inside subgroups, subgroups and regular equivalence groups among subgroups. Comparing with the compress layout algorithm based on cluster and the algorithm for reduse elements of map, SALC algorithm kills bugs of these two algorithms about compress effort isn't so good, correspondences is poor and may lost useful information.A prototype system of information visualization of social network based on ontology which called OIVSSN is designed and developed in this paper. In OIVSSN, the domain ontology model, analysis method, layout algorithm and layout compress algorithm proposed in this paper are realized. Further more, OIVSSN establishs a modular mechanism for analyzing, layout computing and displaying the social network information automatically. OIVSSN has good expansibility.Our research results such as ontology model, analysis methods and algorithms proposed in this paper facilitate the modeling, sharing and reusing of the information visualization technology of social network and are valuable for development of this technology. They have been proved through the typical cases such as terrorist information and advance the researchs of our national 863 projects and natural science fund porject.