Abstract: Abstract: A method using Cited Reference Search in Web of Science to search the matrix of co-citation amounts among journals in the whole database is introduced; the complete co-citation matrix could be got. Then the co-citation ratio of the matrix is calculated and a clustering analysis is made. The co-citation ratio matrix is the standardization of co-citation matrix, which could reduce the data error. After that, the sample journals are randomly selected from 4 subjects in JCR, and an empirical analysis to this method is carried on. The results of clustering analysis fit completely close to the classification in JCR. Further, an analysis on 78 management journals indexed in JCR-SSCI is made, the co-citation ratio matrix is searched and calculated, Netdraw is used to cluster the journals, and analyze the inner knowledge structure and knowledge communication of management science.

Key words: Web of Science, co-citation ratio, journal classification, management, social network analysis