In recent years, profiting from the rapid development of Internet, more and more people can publish information on it, and they also obtain information from it. The web has become a platform which includes information processing, manufacturing, processing and distribution, it provides contents cover all aspects of people's living, study and work, and its speed is continually expanding, so it is a ever-changing source of information. With the development of study on the physical meaning and mathematical properties of complex networks, it was discovered that many networks have a common trait with social network, it called community structure, but this structure is a set which elements are interested in some theme, that to say, the elements have the same properties. Such a structure also exist in a website that have rich resource and user, such as educational website, some users collected together based on similar interests, and form a interest network. Since the community structure also exist the interest network, it also should have a number of sociological characters that the social network embodied, such as:community centrality, community stability and small-world characters. This thesis will use the node similarity method to discover some communities. First of all, we will look the nodes in network as each separate community; next, calculate the similarity between them, merge the two nodes that have largest similarity as a community, and adjust the similarity between the other points and the new community; finally, through calculate the module to find all of the communities.Firstly, we put a research on the method about community modules calculation that be proposed by Newman and Girvan, because its highly subjective and its result will be different under different conditions; Secondly, we will introduce a community discover algorithm based on interest similarity, and achieved very good results.The experimental results show that the scale of communities our algorithm found, is well-proportioned, and most of these communities are relative small. From the stability of community speaking, small-scale community is stable and easy to understand.Using the community discover technique in education website, we can continue find the core theme and the core man, then set a new network environment including planning, learning, self-teaching, in order to make up the disadvantages of the traditional, and lay a foundation for next step to establish a "theme resource community", and provides users with a more professional platform for communication, the community discover technique for educational website must stimulate the development of Teaching information. |