| More and more users have been over a network to share and exchange all kinds of information with the rapid development of computer technology and network. The concept of social network has come to being. Each actor with other actors has more or less relationship in social network, while social network analysis is to establish a model of these relationships and trying to describe the relationship between population structure to study the individual impact between groups or group itself. We could get the users and information that are most popular or most interested in their own through analyzing the relation of characters in social network. It provides great reference value for users to share and exchange information directly. Accordingly, we provide the research of mining and analyzing of characters relation in this paper according to individual behavior in social network.In social network, more and more users adopt the form of Blog to share all kinds of information they get with other people, and others also share and exchange information according to their own interests. Therefore, we take the behavioral characteristics of users in Blog site as the fundamental basis to explore and analyze characters relationship. We call the users that have relation because of some blogs related users, and the themes of the blogs are called related users keywords in this paper. We take the word has the highest weight, which is calculated according to TFIDF algorithm after segmentation, in a Blog as the keyword of the Blog.The paper describes the mining and analysis method of related users in social network, designs and implements related users query system which takes scientists network as experiment target.Firstly, it introduces the research significance and research status of the subject, and then describes basic algorithms and technical theory which are used in the subject research, including MD5, TFIDF, Linux kernel single mechanism multiple IO, network programming and other related technologies. Secondly, it makes requirement analysis and overall design of the related users query system, and describes the algorithms design of key technologies in the system implementation. Including the strategy of duplicate crawl avoid, the storage structure, the extraction of related users and related keywords and the display of result.Finally, it summarizes the work of this research and prospects the future work. |
