Organizational Network Reshaping Problem And Visualization Research

In complex network research,node failure caused by random failure or sabotage is a common phenomenon.Typical networks such as the Internet,communication networks,transport networks are always facing the risks and consequences of this phenomenon.How to reduce the impact efficiently becomes an important subject with practical significance.In the existing studies,scientists always use node similarity to find a successor of the failure node.In many cases,the calculation of node similarity depends on the topological structure of the networks which leads to the problem that the result neglects many inherent properties of the nodes themselves.In this thesis,networks of abundant node properties are modeled as organizational networks.By doing this,we try to solve the problem that when node failure happens which node in the rest of the network will become the successor,how the links will change and how the network will reshape itself.For different organizational networks,network characteristics varies a lot,so both inherent node properties and network topological structure are taken into account.In this premise,the organizational network reshaping algorithm based on node similarity is proposed.After the algorithm is proposed,some effective indicators including weighted network efficiency and network reversion degree are used to measure its effect.Through several simulations,we come to the conclusion that our organizational network reshaping algorithm performs well in node failure problem,it reduces the efficiency loss of the network and also maintain the topological structure at the most extent.Beside all these works,an online visualization platform for researching organizational network reshaping problems are also built as a supporting facility.