The Design And Implementation Of Network Visualization System

In this stage of rapid development of communication and Internet technology, the amount of information showing exponential growth. How to excavate its potentially valuable patterns and structures from these complex information is one of the challenging research topics in the field of data mining.Information visualization technology is an effective means for data mining,which take advantage of the powerful image processing functions of the computer. As an important branch of information visualization, graphical visualization technology abstract information as a node in the graph, and the relationship between the information abstracted as a edge, information is presented as a graphic form.To help users find the forms and characteristics of the patterns and structures through intuitive observation and analysis,give a deep understanding and take full advantage of its implict potential value. Graph layout technology is the core and foundation of graphical visualization techniques.The objectives and tasks of the relationship network visualization system is to develop a common network graph or tree structure information visualization system. The core content of the study is complete a scientific and technological projects of Shanxi Province "mobile customer relationship network visualization system platform".The main contents of this paper include the following aspects:(1)Describes the background, significance and related technical principles of project development. Describes the typical method of relational network visualization and clustering analysis, including flexible layout algorithm and radial tree layout algorithm, similarity measure on node of undirected graph, as well as typical clustering algorithm.(2)According to the "mobile customer relationship network visualization system platform" project task requirements analysis and definition of the system requirements.(3)Relationship network visualization system is designed and implemented using object-oriented approach. The system supports GraphML and GML, as well as user-defined graphical data file.Make it into the data structure that our system can visualize.Completed Force-directed-Layout, radial tree layout, random layout, grid layout and circle layout algorithm and so on. Also integrates a number of typical tree layout.(4)Implement some similarity measurement algorithm such as Manhattan distance, Euclidean distance, cut Chebyshev distance and Minkowski distance on the nodes. On this basis, the system can complete cohesion and framing of the network clustering operation, and visually presenting the results of clustering.(5)The system uses the Java language and NetBeans integrated development environment, you can run a variety of operating system platforms.Experimental results show that the system reached the functional and performance requirements of the mission statement. Completed the functions required of the project. Implement network visualization.