Research On Comparing Journal Performance Across Subject Categories Based On Page Rank

Science Citation Database is the basic tool of academic research,the evaluation of academic publishing, scholars and research institutions based on these data is an important reference tool on the research of resource matching, subscription, professional title appraisal.The Journal Citation Reports(JCR), annually published by the Institute for Scientific Information(ISI), provides many evaluation indicators on academic journals in the sciences,including Impact Factor and Eigenfactor which play an important role in the scientific research work and are the most commonly used journal evaluation index. However, the indicators like Impact Factor and Eigenfactor cannot achieve good results when comparing the journal performance across subject categories, as journals in different categories have great difference in the number of articles, number of citations and the period of publication. Taking account of the development of the interdisciplinary fields and the needs of research institutions, comparing the journal performance across subject categories has very important significance.In this paper, we conduct the algorithm research on comparing the journal performance across subject categories based on the PageRank model. We construct the Personalized PageRank vector which reflects the demand of comparing journal performance across subject categories, and get the evaluation ranking results.The field of journal categories is redefined with vectorization in this paper, and proposing the journal category proportional allocation algorithm to get multivariate quantitative measurement of the journal category attribute makes the journal field attribute reflected more objectively. On the other hand, the algorithm also has timeliness, implying the journal’s preference on different categories and publishing aspects. We proof the convergence and numerical justification of the algorithm and analyze its practical validity with the data experiment.According to the PageRank algorithm, we obtain the rankresults of comparing journal performance across subject categories, with the personalized PageRank vector which is computed by the standardization of the journal category based on the journal category proportional allocation vectors. The standardization overcomes the influence from the difference of journal categories and meets the demand of comparison across disciplines. We conduct lots of data experiments with part of the journals data of the JCR of the year 2008. With the experiment results, we conduct a comparative analysis on Impact Factor, Eigenfactor and Cross-Discipline Eigenfactor which proposed in this paper, and illustrate the rationality and effectiveness of the algorithm in this paper on comparing journal performance across subject categories.