The Research Of Influential Community Search Method On Large Scale Data Graph

Community discovery problem can be applied to biological networks,traffic networks,infectious disease prevention and control network and social network,used to find a specific feature,a set of closely related entities.Community influence is an objective measure of the importance of the collection of objects in the network.With the rapid development of network technology,a large scale of data has appeared in the practical application.There are a lot of communities on a big picture,but in reality,users are only concerned about the closely linked,influential community.In this paper,we study the problem of query top-r influence community,which is used to return the highest r communities from the given undirected graph.Specific contents are as follows.Firstly,in the aspect of index building,we propose to organize the relationship among the communities by index tree,In order to construct the index tree quickly,firstly we calculate the maximum k value of the k-core on the original graph.Then,each node of the graph has a corresponding father node mark in the index tree.After calculating the node in an index tree,set its parent node in the tree.When all nodes in the graph are processed,the index tree is constructed.Compared with the existing methods,it reduces the time,and improves the construction efficiency.Secondly,in the aspect of community query,we have redefined the top-r influence community.We use the weighted average of the community nodes as a measure of community influence.This avoids the sudden decline of community influence when a new user to join the community,and the top-r results are more practical.At the same time,this paper proposes a new top-r influence community query algorithm,which is based on the new index structure,fast output all top-r communities and their influence.Finally,in the experimental stage,we compare from three aspects,the index construction time,query time and index size on 6 real datasets and 6 artificial data sets,validated the efficiency of the proposed method in this paper.