面向研究社区的论文级评估方法.pdfVIP

面向研究社区的论文级评估方法 方婧，刘宇** （大连理工大学软件学院，辽宁 大连 116621 ） 5 摘要：随着科技文章的迅速增长，如何从大量论文中选择高质量的文章变得尤其重要。基于 这种情况，如何从大量论文中选择高质量的文献对研究者是越来越重要。在这种情况下，论 文级评估方法被提出用于去衡量单篇文献的质量。社会性书签量被许多研究人员认为是当前 最好的对文章前期质量的评估；引用量随着文献发表年限的增加能够很好地反映文章质量。 但是当前的研究还没有关于如何利用两者的优点对文章质量进行评价。本文提出的面向研究 10 社区的论文级评估方法，不仅综合了社会性书签量、引用量、作者和文章所在期刊对文章质 量的影响，还充分利用了研究社区中存在的文献列表。实验结果表明，该方法是足够有效和 健壮地评估文献质量。 关键词：计算机应用技术；论文级评估方法；社会性书签量；引用量 中图分类号：TP39 15 Article Level Metric in a Research Community FANG Jing, LIU Yu (Software School, Dalian University of Technology, Liaoning Dalian 116621) Abstract: With the rapid increase of research papers , it is important for researchers to choose the 20 high-quality papers from a large number of papers. In this context, article level metrics are proposed to evaluate individual paper. Many researchers thinks the number of social bookmarks is the best method of evaluating the quality of recently published papers among these metrics . Moreover, the number of citations best reflect the quality of papers with time passed by. However , the current research does not study how to systematically the merits of these two methods to 25 assess papers . In this paper , a novel article level metric called ALM_RC is proposed. It not only systemactically integrated the impact of researchers , journals , social bookmarks and citations , but also makes use of articles listing in the research community . Experimental results o