The System Clustering Of Web Impact Evaluation For Journal

Since the 1990 s, a variety of mainstream journal evaluation agency undertake the work of journal academic impact evaluation, for more than 6000 kinds of academic journals comprehensive evaluation, related scholars have initially established a set of evaluation systems which combined quantitative and qualitative evaluation. With the rapid development of information technology, the Internet has been integrated into all aspects of human life, it’s no longer simply to stay in the traditional printed journal which to conduct research, more and more people use mobile devices, through the full text database to get the latest research,and at the same time the demand of the network influence of journal evaluation is highlighted increasingly.Using system clustering method in this paper, for journals of library and information science which are on the network database to classify, get the classification about the web impaction of journals of library and information science. System clustering method can classify according to the similarity in nature though there are differences on their own individual data, using this method to clustering the statistics, you can clearly get the grade distribution of journals of library and information after the results dealing with the system clustering method.In this paper, as the main data source libraries are " China Academic Journal Network Publishing Database " and "Chinese citation database", data of the research object base on the 33 kinds of library and information science journals, statistical data include: number of papers, cited quantity, download quantity, H-index, cited quantity of each, download quantity of each and web diffusion coefficient. Using SPSS20.0 software for data clustering analysis, get ice pillar diagram and dendrogram, two figures can objectively demonstrate the process of clustering and clustering results. According to set a threshold for classifying the clustering results, to level to distinguish all kinds of journals base on comprehensive coefficient. Contrast with The evaluation results of the traditional journals and the clustering results, have a depth discussion about web impact factors, providing reference and ideas for future study.