Demonstration Research And Analysis On Directed Subgroup Network Of University Websites

With the continuous development of computer technology, people can search through the network of mass WWW network database, and can study contains millions of Web pages and 100 million hyperlinks on the Web links in the WWW complex system. In some of these areas on topics such as structure of the application of information retrieval; WWW links rich and complex information structure hidden inside what characteristics and through what mechanism for self-organizing into a complex system, attracting ever more and more Chinese and foreign scholars'enthusiasm.In order to study the characteristics of such a large-scale system and the rule, a new theoretical research methods - complex network theory, be greatly to the development and application. This theory is that WWW network nodes and edges of the complex network structure, to study the network topology characteristics and dynamic characteristics. Scholars at home and abroad, mainly from the perspective of overall situation of the whole WWW network of extensive study, and very few people on the WWW of the local network to do in-depth research. Therefore, in this paper basis of the results of previous studies, we research from the perspective of local colleges and universities portal consisting of the complex subgroup network.To sum up, the main works of this paper are as follows:First, this paper designs automatically search algorithm for large-scale web page. It combines the hierarchical search algorithm, breadth-first search algorithm and related priorities. The use of the algorithm on some colleges and universities to automatically search portal site, the establishment of colleges and universities have to Subgroups network database. These algorithms use, saving a huge amount of time to search the network and solved the problem of the collection of raw data.Then, according to Network Database Construction of a subgroup network to the web page abstract map of the nodes, the hyperlinks abstract connections between nodes, and studied the subgroup of network topology such as vertex degree distribution, cluster coefficient, betweenness and joint degree distribution. According the result of statistical analysis, we found that this kind of subgroup network has some similar properties. All of the networks, for example, have the features of scale-free and small-world effect. In addition, it was also found point-two showing a distribution of power law distribution and the joint distribution of show zonal distribution characteristics.Finally, according to University subgroup network link topology characteristics and the general rules of web design, we have proposed a directed network model with growth, preferential attachment and rewired mechanism. Through computer simulation we found that the University network model and the actual Subgroups with network topology features are more consistent. The mechanism of the evolution model is more in line with colleges and universities subgroup network growth patterns.