Research Of Complex Networks Community Structure Mining And Its Application

Many real world systems exist with the network form, such as interpersonal networks, internet and so on. With the In-depth study for the physical meaning and mathematical properties of the network, we find that many real networks have a common nature which is community structure. Complex networks community structure mining has great theoretical significance on the analysis of complex network topology, understand the function of complex networks, find the hidden rule in complex networks and predict the behavior of complex networks. And also, it has a broad application prospects. Now it has been used in terrorist identification, organizational structure management of social network analysis, metabolic network analysis and other fields.There are already many community structure mining algorithms. But almost all these algorithms have the problem of high time complexity or space complexity and can only handle the general complex networks which only contains "positive relationship", signed network is that contain both positive relationship and negative relationship which promotes the description capacity of complex network. This paper proposes a algorithm for community structure mining based on greedy algorithm. Its time complexity is relatively small and it is applicable to the signed network community structure mining. Experimental results compared with the GN algorithm verified its effectiveness and feasibility.As the field of complex networks research emerges continuously, how to apply these results has become a new hotspot. This paper does exploratory research on the application of complex network theory.This paper mainly completed the following work: (1)Implement"GN"algorithm, then apply it on Zachary network and analyze its performance.(2)For the weak point of"GN"algorithm, proposed an algorithm for community structure mining. Compared with"GN"algorithm, the algorithm proposed shows high performance,(3)Extend the proposed algorithm to solve the signed network community structure mining problem.(4)Build the physical model of power dispatching network and the corresponding adjacency matrix; use the proposed algorithm to get the electric power dispatching network subnet topology.(5) Based on the small world network theory, does the network topology analysis for the three phase rectifier circuit, build the small world network model under a variety of short circuit and open circuit, then calculate characteristic parameters, summarize its network characteristics and features, and explore their laws, lay the foundation for further study.