The Law Of Paper Collaboration Of Computer In China Based On Social Network Analysis

If we consider an author as a vertice, there is an edge connected between themwhen they have authored a paper together.Essentially speaking, paper collaborationnetwork is a special type of social network.M.E.J.Newman, a famous scholar, thinksthat paper collaboration network is a kind of relation network, which is formedbecause of exchange and collaboration between researchers in the process of scientificresearch.Taking20kinds of core journals of computer fields between2000and2011inChina as data sources, computer field paper collaboration network between2000and2011in China has been made;Paper collaboration network for each journal between2000and2011in China have been made;Computer science and software technicaljournals paper collaboration network,computer image and graphics journals papercollaboration network,computer automation and control journals paper collaborationnetwork,sensor theory and application journals paper collaboration network have beenmade. Based on social network analysis, features of paper collaboration networksmentioned above are analysed from individual and integrity attribute aspect, and someconclusion and law are made as follows:(1) Paper collaboration network is a scale-free network, which has an obviouscharacteristic of small world and core nodes with large degree value and largebetweenness value.(2) Paper collaboration network is multi-core network structure.Networks are notfully connected network. Each of these networks has some cliques, and most cliquesare national key labs or research institutes.(3) Paper collaboration network can be divided into kernel network andperiphery network. Most of the core elements come from kernel network; someperiphery network elements have large degree and large betweenness, which provesthat some periphery network elements play a certain role in the network connection.(4) Generally speaking, core elements are reserachers with large degree and largebetweenness. Order of degree and betweenness is not identical in a network.Mostnodes with large betweenness and small degree are crossdisciplinary andinterdisciplinary researchers. (5) Computer science and software technical journals paper collaborationnetwork has a large mean distance and a small mean clustering coefficient, which hasa small-world effect, but less obvious.(6) There are some discrete nodes in every network, and most of these discretenodes are freshmen of paper collaboration network.Finally, a clique network constrution study has been done and some suggestionsare proposed.