The Research Of Digging Community Structure And Mobility Patterns Analysis Based On The Complex Network Theory

In recent years, the development of complex networks theory provide theoretical models and research framework for people to understand networks in the real world such as Internet, Biological networks, WWW, Blog networks, etc. Recent studies suggest that a lot of real world networks have properties like small-world effect, scale-free, gathering and community structured. Much recent work in application of complex network theory and Web data mining has been devoted to mining meaningful information from the large real-network, community structure analysis and network behavior prediction using behavior data extract from real-world networks.Community structure is an important characteristic of complex networks. Mining community structure reveals the internal structure of complex networks, which affect the evolution and dynamic behavior of the networks. So far, researches on social networks are mostly based on static or evolving network models. In a static model, the real-world network is modeled as a time-independent network; while in an evolving model, works are focused on the new nodes and links. Many significant findings like small-world and scale-free fit networks like www and citation networks well. Nevertheless, these methods seem not efficient enough when applied to social networks. As a typical social network, human behavior can be efficiently modeled by constructing proper rules and network.The research of digging community structure and mobility patterns analysis in behavior network can be beneficial to more precise behavior prediction and solutions to improving efficiency in life; also it is the foundation and basis of modeling and analysis the social network.In this thesis, analysis of taxi data from Bus Company of Shanghai and interrelationship between blogs has been approached with complex network theory. The outline of this paper is as follows:1. Attain the dynamic trajectories of the taxis from the raw data from Bus Company of Shanghai, establish a new perspective to understand the small-world effect, scale-free, aggregation properties and the community structural characteristics of complex network;2. Compare the behavior features of taxis and human from the perspective of complex networks including the trajectory, the probability distribution of step, the cyclotron radius growth over time, time and space characteristics, anisotropy of activities, etc. Although the taxis actives is different from human behavior, but also follow the same scale-free features of complex network.The study provides a basis contact for building dynamic network model, it also has some meaning to forecast the functional areas of Shanghai with the data and optimize the resource allocation and interest income of taxis;3. Visualize and verify interrelationships between blogs and find the community structure in blog space. Application the massive data processing method divides the blog space into seven community structure by using a combination of hierarchical and modular approach and calculates the value Q of the corresponding community structure. Verify the blog space has the basic characteristics of complex networks.