Research And Implementation Of Visual Analysis Technology For Community Structure In Social Networks

"Network" exists everywhere in our lives:friendship in our social circle is a network,the relationship between the web pages is a network,and the international trade among countries in our real world is also a network.A deep analysis of the network data will help people to have a better understanding of the knowledge and wisdom that hidden under the network structure.However,today,traditional characters and tabular forms could no longer meet people's growing need of internet data analysis and management.Network data visualization is a very important part of Information Visualization Technology.It will make full use of human's perception system to help users understand the network data structure,and it makes the network data to show us in the form of graphics,then can dig deeper information from it.Based on the deep study of social Network Analysis Technology and Visual Analysis Technology,this thesis analyzes and compares the layout effects of various kinds of force-directed algorithms,and puts forward a layout algorithm which can show the information of community structure information.This algorithm can help users to carry out further community division and key nodes analysis.Finally,the author designs a prototype system to achieve a social network in the community structure of the visualization analysis.Specifically,the main tasks in my thesis are as follows:1.On the basis of studying the theory of social network analysis and visual analysis,this paper puts forward a visual analysis model which is suitable for community structure in social network.2.In order to realize the layout effect which can reflect the network structure,this paper proposes a set of evaluation system based on the social network characteristics,which consists of aesthetic rules,clustering effect and running efficiency.Finally,the design experiment is used to evaluate the two classical force guidance algorithms-FR algorithm and KK algorithm.3.Aiming at the drawbacks of the classical force guidance algorithm which can not show the information of the community structure in the social network,an improved layout algorithm with better clustering effect is proposed.This algorithm improves the repulsive force and gravitational force formula in the traditional force-directed algorithm,and introduces the concept of in credential similarity,and achieves to display its structure without dividing communities in advance.Experiments show that the improved algorithm has got a better performance than the traditional force-directed algorithm in demonstrating the clustering of social networks.It is because the improved algorithm can effectively show the community structure in the network that can helps the users to do an intuitive interactive community division.4.Design and implement a visual analysis system for the community structure in the social network,and realize the visualization and interactive functions of the network by assisting users such as network layout,community division and key node analysis.