Decipher the Hidden Secrets Behind Networks: Analyze the Influence Based on the Co-author Network

摘要

In a scientific research network, to determine the influence of an academic research, a key is to build and evaluate properties of its cited references or co-author networks. To analyze 18,000 lines of original data from Problem C given in 2014 MCM/ICM (Mathematical Contest In Modeling /Interdisciplinary Contest In Modeling), the Pajek software was used with obtaining a result that correlations among those 511 authors are not strong. Then, to get each author's ranking of academic influence, the AuthorRank algorithm was introduced and adopted. In requirement 3, to evaluate the given 16 essays, five indexes were chosen: Cited rate, H5-index, IF (impact factor), H-index and I10-index. Using the entropy method and Matlab software, the comprehensive values of each paper for their influence are obtained. Finally, to further verify the feasibility of the AuthorRank algorithm, 5,000 lines of data concerned with relationships in DouBan is gathered and analyzed in our study.