The Design And Implementation Of Social Network Data Acquisition And Structure Analysis System

The coming of Web2.0era makes the Internet technology towards a more humanied way, social software such as Twitter, Facebook, Micro-blog, Friends network, Renren and so on, also rise subsequently and develop rapidly. At present, people’s daily communication activities basically carry out on the platforms provided by these social software. People use these social softwares as mediums to exchange information with each other for some purposes, resulting in relation networks, this network structure is consist of people, and the relationship between people, called the social network. Two structural elements of social network are node and edge, node is generally an individual, edge is the relationship between people. Conform to the needs and development of science and technology, resulting in cooperation network, which is the product of research collaboration, and which is the social network among scholars, and co-author network is a network of cooperation network produced by co-authored papers by scholars, it is a social network between co-authors. The study object of this dissertation is the relationship network between micro-blog users and co-author network, which are the representative of the social network, the former is a directed network, the latter is an undirected network.Since the concept of social networks put forward from sociology, it caused the wide attention of scholars both at home and abroad, so far, the study of social networks boom has yet to recede. Network data acquisition is the primary problem to solve on social networks research, however, most of the existing research on social networks, the source of the network data is public data sets, or simulation of network data sets, thus in some ways, they can not accurately reflect the actual circumstances of the social networks structure, therefore, getting the real data of social network structure from the Internet shows especially important, and makes the research achievements of social networks more practical significance. In this dissertation, we design a social network data acquisition and structure analysis system and this system has realized the real data acquisition from Sina micro-blog system and DBLP database which reflect the real relationship between Sina micro-blog users and the real relationship between co-authors respectively.Social network analysis method and complex network analysis method are widely recognized both at home and abroad which are used for social network structure analysis. For scientific research co-author network, analyzing its network structure, plays an important role in promoting the continued development of the scientific research cooperation, and predicting a development direction in the field. For micro-blog users network, analyzing its network structure has important significance in the market operation, user recommendation. The system is designed and implemented in this dissertation by using the method of the role analysis of social network analysis method to analyze scientific research co-author network structure, the opinion leaders and structure holes are study objects, and the complex network analysis method to study the network topology characteristics of the relation network between Sina micro-blog users.The dissertation designs and implements a system of social network data acquisition and structure analysis, the main work is as follows:1. This dissertation describes some concepts and techniques which used in the process of designing and realizing this system.2. The system has designed and implemented the capability to implement Sina micro-blog users network data acquisition and structure analysis, the system can be done to get the real relationship data between users from Sina micro-blog, the data de-noising processing, and generates the graph of relationship network structure, then, use of complex network analysis method to analyze characteristics of the network topology.3. The system has designed and implemented capability to implement co-author network data acquisition and structure analysis, the system can be done to get to the co-author network data from the papers recorded by the DBLP database about four levels of conferences, which theme is" data mining ", process data to generate network structure graph, and detect the100top of structure holes and opinion leaders. 4. Take the100top of structural holes and opinion leaders as the study object, compare and analyze them by the number of papers, citation numbers, H-index and G-index which are measurable indicators for measuring scholars’ academic achievement.