期刊引文网络影响测度研究

Measuring the Relative Impact of Journals in Their Citation Net

利用Goosle的PageRank原理进行期刊引文分析，提出期刊在引文网络中的影响力测度指标——引文网络影响力序位（Journal Impact Rank in Citation Net，Impact Rank或IR）。通过对118种生物学领域的期刊进行期刊引文网络影响力测度，并将IR结果与JCR提供的影响因子（Impact Factor，IF）值进行统计学分析以考察二者的相关性和差异性。结果表明，IR值与IF值的相关性较弱，其差异性具有统计学意义。分析其原因，IR考虑了引证期刊的权重和期刊间的相互影响，更适于反映期刊在其相关学科或领域的引文网络中的相对影响力；IF值因其实质上是期刊论文篇均被引频次，其计算不考虑期刊之间的相互联系和引证期刊的权威性，因而更适用于期刊自身的纵向评价；IR与IF从两种不同角度评价期刊影响力，可互为补充。

We propose a new indicator of journals impact basing on the PageRank algorithm for Google,Joumal Impact Rank in Citation Net, for short, Impact Rank or IR. IR was used to measure the impact of 118 biological journals in their interacting citation net. The correlation and difference between IR and Impact Factor in Journal Citation Report （IF） were statistically analyzed. We testified that IR was weakly related to IF but statistically different from IF. We conclude as following ： IR is of more potency in signifying relative journal impact in citation net of a given discipline/field than IF by taking into account the weight of citing johrnals and the interactivity of journals. IF can be used to longitudinally evaluate an isolated journal itself, because it is virtually the average citation frequency per paper in a journal, regardless of the interactivity between journals and the authority of citing journals. IR and IF are two mutually complementary indexes for evaluating the academic impact of journals from two distinct points of view.