The Research On Visualization System Of Social Networks Based On Prefuse

Scientific research found that most of the real-world systems show the features of complex networks, such as a heavy tail in the degree distribution, a high clustering coefficient, assortativity or disassortativity among vertices, community structure, and hierarchical structure. Discovering those features of complex networks is very challenging and has important practical application value. In recent years, by combining the visualization and clustering analysis technologies, a new hot research branch about complex networks has emerged.The graphical visualization technology, which represents network data sets in graphical forms on the computer screen, has been considered as one of the effective means of detecting the complex networks due to its inductrvity, interactivity and visual capability. Combined with the clustering analysis technologies, graphical visualization can assist researchers to conveniently analyze, observe and capture valuable organizational patterns and structural characteristics hidden in the networks. However, few practical commercial tools with both the network visualization and clustering analysis functions are available at present. Consequently, the development of such a visualization system has important practical significance and application value.The visualization system of complex networks presented in the paper is constructed based on Prefuse library licensed under the terms of BSD license. The software is platform-independent, easy to expand and convenient to use, and several typical graphical layouts for visualizing semi-structured data sets have been integrated into it.The main research contents and technical characteristics related to the software development are the follows:(1) A detailed software requirements specification of its functions and performance is delivered, the system architecture, some key algorithms and data structure etc., are addressed.(2) The system is written in Java programming language and developed in the object-oriented technologies under the popular integrated development environment NetBeans.(3) In addition to the self-defined type, the system also supports the common standard graphical data input formats, such as GraphML and XML.(4) Several typical graphical layout algorithms, such as the force-directed, space-optimized tree, radial tree, node-linked, random layout, and grid layout etc., are implemented and integrated.(5) The implanted similarity measurement models of network vertices based on the Manhattan distance, Euclidean distance, Chebyshev distance and Minkowski distance, On the basis of the above algorithm, to network node data obtained, implant the clustering operation of different clustering algorithms, and the clustering result visualization.(6) The visual properties of graphical objects, Such as their sizes, colors and shapes can be interactively configured, and any changes will cause the graphs to be regenerated and refreshed immediately.The experimental results show that the predefined fundamental functional and non-functional requirements are achieved, the interactive visualization of complex networks can be performed easily, and the implementtation has laid a solid foundation for the further development of the system in the future.