摘要
本文构建了基于位置的共被引分析方法框架,并通过实证探究,得到如下结论:引文在全文中的分布极不均匀,大部分引用都集中在文章的前面部分,且引用经常成簇出现,不同类型的文献引用分布不同;文献对的相似性不但与共被引频次有关,也与共被引层次有关,传统的共被引分析结果掩盖了非文章层次的效果;各共被引层次的共被引发生的可能性随共被引频次的增加而减少,共被引频次与各共被引层次的共被引里负线性相关关系;基于位置的共被引分析方法能够提升共被引聚类的效果,有利于进行深层次研究和评价。
In this paper, we constructed the location-based co-citation analysis framework and performed an empirical study. Conclusions show that ： Firstly, the citation location distribution in the full text is uneven, most of the references are cited in the first part and they often appear in clusters. The citation location distribution of different kinds of literatures is different. Secondly, similarity of literatures is not only related to times co-cited, but also related to the level of co-citation, the results of traditional co-citation analysis cover effects of the non-article level. Thirdly, the probability of occurrence of each level of co-citation decrease along with the increase of citations times co-cited. Times co-cited is significantly negatively correlated with the level of co-citation. At last, the location-based co-citation analysis can improve the accuracy of co-citation cluster, and it is beneficial to carry out in-depth research and evaluation.
出处
《情报学报》
CSSCI
北大核心
2016年第5期492-500,共9页
Journal of the China Society for Scientific and Technical Information
基金
教育部人文社会科学基金项目"馆藏数字资源语义化深度聚合的理论与关键技术研究"(13YJA870023)研究成果之一
关键词
位置
共被引分析
共被引层次
共被引距离
实证研究
location, co-citation analysis, level of co-citation, proximity of co-citation, empirical research
