基于互信息强度构建标签概念层次结构方法的探究-情报杂志.pdfVIP

第33卷摇 第12期 情摇 报摇 杂摇 志 Vol.33摇 No.12 2014年12月 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 JOURNAL OF INTELLIGENCE摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 摇 Dec. 摇 2014 基于互信息强度构建标签概念层次结构方法的探究* 江雪琴摇 张志平摇 李琳娜 (中国科学技术信息研究所摇 北京摇 100038) 摘摇 要摇 大众分类、社会化标注等个性化标签系统对网络资源的组织优势很明显,但存在标签多样和概念结构模糊 的缺陷。 文章以“豆瓣冶读书标签为例,探讨自由标签系统概念上下位关系的识别方法。 主要是先对标签进行聚类, 将其分成若干个内部结构联系紧密的类簇,然后以互信息强度,具体判别上下位关系,由此构建标签层次结构。 实验 表明构建完成的概念层次结构图,能够比较准确地识别标签间的概念关系,为用户提供良好的标签导航和浏览机制。 关键词摇 标签摇 层次关系识别摇 聚类摇 互信息强度 中图分类号摇 TP311摇 摇 摇 摇 摇 摇 摇 摇 文献标识码摇 A摇 摇 摇 摇 摇 摇 文章编号1002-1965(2014)12-0165-05 DOI摇 10.3969/j.issn.1002-1965.2014.12.030 Research on the Method of Tag Concept Hierarchy Construction Based on the Strength of Mutual Information Jiang Xueqin摇 Zhang Zhiping摇 Li Linna (Institute of Scientific and Technical Information of China,Beijing摇 100038) Abstract摇 Personalized tagging systems,suchasfolksonomyand socialtagging,playanimportantroleintheorganization ofnetworkre鄄 sources. However,there are the defectsof the diversity of thetagsusedbyusersandtheunclear conceptual structure. In thispaper,tags from “Douban冶 are chosen asthe dataset in our experiment. At first,tags are divided into several clusters with closely-linked internal structure. Then,the hyponymy relationsof tagsare recognized based on the strength of mutual information. The resultsdemonstrate that the concept hierarchical relationship among tagscan accuratelyrecognizethehyponymyrelationssothat it canhelp thetagging systemsto provide userswith good tag navigation and browsing mechanisms. Key words摇 tag摇 hierarchical relationship recognition摇 clustering摇 the strength of mutual information 深受研究者们的关注。 0摇 引摇 言