Research On The Method Of Discovering Computer-Communication Network Organization Structure

With the development of information technology,the communication based on computer has attracted the attentions of all sectors of society.In companies or groups,the computer-based communication such as e-mail and instant messaging has become one of the most popular communication means.Thus,the research about computercommunication network has gradually become another focus of social network analysis.In most computer-communication networks,the importance of each node is not equal,and the nodes in network has been divided into two parts,namely important nodes and ordinary nodes.We can speed up the information dissemination or quickly destroy the entire network communication by manipulate the important nodes of whole network.Therefore,it is significant to find out important nodes of computer-communication network.In addition,computer-communication network can also reflect the reality organization structure of group or company.This organization structure is always hierarchical,and nodes at different levels will have different position,the node at the highest level is general leader of whole group but the nodes at the lowest level are ordinary members,and there is a leadership between nodes at adjacent levels,that is,nodes at upper level manage nodes at lower level.Therefore,this paper will focus on two aspects,namely important nodes selection from computer-communication network and organization structure discovery of computer-communication network.The main work is as follows:1)In order to select important nodes from computer-communication network,this paper puts forward a type of important nodes selection algorithm which is based on scoring model.This algorithm takes nodes' communication behavior and topology structure into account,and then establish a scoring model according to this two type of features.Thus,each node will has a score through this scoring model,the higher score of node,the more likely to be an important node.After each nodes' score is obtained,important nodes of the whole communication network would be found according to nodes' score distribution.2)On the basis of important node selection algorithm,this paper has also put forward a type of organization structure discovery algorithm.This algorithm will call hierarchical clustering algorithm to obtain hierarchical community structure of communication network,and find out the leader node of each community by computing nodes' PR value and importance score.Then the final organization structure could be found by integrating hierarchical community structure and community leader nodes.At the end of this paper,three computer-communication networks are used to verify the effectiveness of important node selection algorithm and organization structure discovery algorithm which is proposed at this paper.It is show that our important node selection method and organizational structure discovery algorithm can find out most of the important nodes and correctly discover organization structure of this three communication networks.