基于文本相似度分析的期刊引文有效性识别研究  

Research on Identification of Journal Citation Validity Based on Text Similarity Analysis

作　　者：鞠秀芳 Ju Xiufang(Institute for Chinese Social Sciences Research and Assessment,Nanjing University,Nanjing 210093,China)

机构地区：[1]南京大学中国社会科学研究评价中心,江苏南京210093

出　　处：《现代情报》2018年第11期14-17,共4页Journal of Modern Information

基　　金：国家社会科学基金项目"基于引文分析的期刊非正常引用行为研究"(项目编号:13CTQ034)

摘　　要：真实性、准确性、直接性与完整性是引用参考文献的首要准则,然而种种研究表明,当前的研究论著中引而不注、过度引用、模糊标注甚至虚假引用等不当引用行为日渐增多,这极大地影响了科学研究领域的学术风气,给读者阅读、期刊审稿及成果评定等工作带来了许多不便。本文利用文本相似度算法建立期刊引文有效性识别方法,试图从海量期刊引文数据中识别出期刊引文的真实有效性。实验表明,本文的期刊引文有效性识别方法在引文有效性方面实现了较好的识别效果,可为虚假引文的识别工作提供可靠的依据,从而为编辑人员发现、修正虚假引用问题提供帮助,彰显科学研究的严谨务实精神。Truth,accuracy,directness and completeness are the primary criteria for citing references.However,various studies have shown that miscitations such as omission,excessive citation,fuzzy labeling and even false citation in current research works are becoming more and more normal,which have greatly affected the academic atmosphere in the field of scientific research.People s reading,periodical review and evaluation of results have brought a lot of inconvenience.In this paper,the text similarity algorithm was used to establish a method to identify the validity of periodical citations,trying to identify the true validity of periodical citations from a large number of periodical citation data.Experiments showed that this method achieved a good recognition effect on Citation validity,and provided a reliable basis for the identification of false citations,thus providing help for editors to find and correct false citations,and highlighting the rigorous and pragmatic spirit of scientific research.

关 键 词：文本相似度 虚假引文 引文有效性 向量空间模型 

分 类 号：G250.252[文化科学—图书馆学]