Research Of Scientific Papers Screening Method Based On Complex Network Of Different Mechanisms

Preservation and dissemination of academic knowledge is human beings s most important vector, the level of its quality and value whether the judge has been the hotspot. Poor quality paper, especially the existence of inauthentic papers is not conducive to the dissemination of knowledge, but also has an extremely negative impact, it will not only take up valuable resources but also more seriously mislead views of others. In order to use the computer automatically screening papers and save researchers s valuable time, this was used in the formation mechanism of complex network theory on the authenticity of the paper part of the screening of the exploratory work.This article first describes several indicators signifying the structure of complex networks, including the subsequent use of the average shortest path, clustering coefficient and degree distribution, and analysis complex network generated by most popular types of network formation mechanism models such as generate fomulated network by coupled model, generate random network by ER model generate small-world network by WS model and generate free network by BA model. Subsequently, this calculation proved by analyzing the network generated by the network generated in the current network size paper under the same terms with them under a large number of the scale reflected the structure of the network nature, thus we campare the network generated by the network generator on behalf of the random network, small world network, scale-free networks with the same term paper size is the network on the network structure of the network.Then we explain the differences.In the end,we use the difference to design a screening method of the authenticity of paper and discrible the result.In the Case analysis of the paper, the paper uses the method to screen the collection of 400 true papers and inauthentic paper. Screening results show that the method can identify the authenticity of paper, and found in the screening process really thesis papers closer than the pseudo-small-world networks, but also different from the random network.