计算机应用研究2017，Vol.34Issue(7)：1966-1970,5.DOI:10.3969/j.issn.1001-3695.2017.07.010
基于优化密度的耦合空间LDA文本聚类算法研究
Coupling space LDA text clustering algorithm research based on optimizing density
摘要
Abstract
Aiming at the problem that traditional vector space model to calculate the similarity in text representation only use statistic the frequency of the word to represent text and to the high-dimensional effect decreased of text data clustering,the paper proposed a coupling space LDA text clustering algorithm based on optimizing density.Linear fusion coupling space model and LDA theme model for computing text similarity,and optimized the issue of threshold of sensitive,the radius of threshold corresponding to the different density area.Experimental results show that,comparing with the improved DBSCAN text clustering algorithms and R-DBSCAN text clustering algorithm,the proposed algorithm performs higher accuracy and better clustering effect in text clustering.关键词
文本聚类/耦合空间模型/LDA主题模型/密度/阈值Key words
text clustering/coupling space model/LDA theme model/density/threshold分类
信息技术与安全科学引用本文复制引用
邢长征,赵全颖,王伟,王星..基于优化密度的耦合空间LDA文本聚类算法研究[J].计算机应用研究,2017,34(7):1966-1970,5.基金项目
国家自然科学基金资助项目(61402212) （61402212）
辽宁省高等学校杰出青年学者成长计划资助项目(LJQ2015045) （LJQ2015045）
辽宁省自然科学基金资助项目(2015020098) （2015020098）
辽宁省教育厅城市研究院一般项目(LJCL008) （LJCL008）
