Information Mining Based On Complex Network Theory

With the development of social informatization and the increase of mobile communication network data,the number of network fault warning logs is increasing day by day.Therefore,it is difficult for communication operators to mine the relevant information efficiently and accurately from the network logs with huge data volume and complicated connection relationship between network nodes.Moreover,if the maintenance and fault handling of the communication network are not timely,the user experience will be affected.The factors affecting the robustness of communication networks are also to be studied.Based on the above background,this paper conducts research on information mining based on complex network theory to solve the existing problems.The main research work is as follows:1.In consideration of the large scale of communication networks,complex relationship of network nodes and the great volume of fault warning logs,it is difficult to mine the relevant information accurately and efficiently.Based on the complex network theory,the undirected graph model of the communication network is built.On the basis of graph model,the network size of a single mining process is reduced by community detection.Topological connections are used to improve the sequence mining algorithm.Simulation results show that the proposed algorithm can effectively improve the accuracy of mining results,shorten the operation time and improve the efficiency of the algorithm.2.In view of the phenomenon that nodes in real networks like communication networks tend to have more connections with others in close distance,the problems in the business scenario are abstracted into the theoretical model,and the influence of spatiality on the network robustness in the complex networks is studied.Actual links and spatial connections are generated in the grid to simulate the connections in the real network.Based on the probability model of spatial linkage generation,experiments are designed to analyze the influence of spatial embedding strength,average connection length and average degree on network robustness.An algorithm is proposed to change the length of connections without changing the degree value of nodes.The simulation results show that the distribution of nodes in geographic space shows that the nodes have more links with other nodes in close distance,which makes the network have better robustness.The same conclusion can be drawn from the expansion experiment of the sparse grid in this scheme,and it is found that the sparse grid is more fragile when compared with the original model.