构件标识潜在语义分析与模糊聚类方法研究
|
摘要
针对如何为构件检索提供更合理的候选构件集问题,设计了一种基于标识潜在语义分析的模糊聚类方法 (TL-SAF)。首先设计了标识提取算法从构件描述文档中提取出构件的标识,然后应用潜在语义分析对标识进行降维并提取潜在语义关系,最后使用模糊聚类的思想对构件进行聚类。TLSAF聚类摆脱了传统聚类的硬划分模式,使构件可以分别隶属于不同的簇,对构件检索将能够提供更好的支持。通过在原型构件库中应用TLSAF对本文方法的可行性与有效性进行了验证。
For providing more reasonable sets of candidate components in components retrieval,a tags latent semantic analysis based fuzzy clustering(TLSAF) approach is proposed.First,an algorithm of Tags extraction is designed to extract the Tags from description documents of components,and then,the LSA is used to reduce dimension and analyze the latent semantic relations of Tags,and the fuzzy clustering is used in the last stage of TLSAF.The TLSAF clustering,breaking the rigid division mode with permitting a component to be classified into more than one cluster,can provide better support for components retrieval than traditional components clustering.Through the application in prototype component repository,the feasibility and efficiency of TLSAF clustering is validated.
For providing more reasonable sets of candidate components in components retrieval,a tags latent semantic analysis based fuzzy clustering(TLSAF) approach is proposed.First,an algorithm of Tags extraction is designed to extract the Tags from description documents of components,and then,the LSA is used to reduce dimension and analyze the latent semantic relations of Tags,and the fuzzy clustering is used in the last stage of TLSAF.The TLSAF clustering,breaking the rigid division mode with permitting a component to be classified into more than one cluster,can provide better support for components retrieval than traditional components clustering.Through the application in prototype component repository,the feasibility and efficiency of TLSAF clustering is validated.
引文
[1]MA Hua.Optimization method on component composition fordomain feature clustering[J].Computer Engineering and Ap-plications,2009,45(21):197-200(in Chinese).[马华.面向领域特征聚类的构件组装优化方法[J].计算机工程与应用,2009,45(21):197-200.]
[2]REN Yaopeng,CHEN Lichao,ZHANG Yingjun,et al.Im-proved method of component clustering based on latent semanticanalysis[J].Computer Engineering,2011,37(4):67-69(inChinese).[任姚鹏,陈立潮,张英俊,等.基于潜在语义分析的构件聚类改进方法[J].计算机工程,2011,37(4):67-69.]
[3]ZHANG Yingjun,REN Yaopeng,CHEN Lichao,et al.Com-ponent clustering algorithm based on semantic similarity and op-timization[J].Computer Engineering&Design,2010,31(11):2531-2535(in Chinese).[张英俊,任姚鹏,陈立潮,等.基于语义相似度与优化的构件聚类算法[J].计算机工程与设计,2010,31(11):2531-2535.]
[4]Kamalraj R,Rajiv Kannan A,Ranjani P.Stability-based com-ponent clustering for designing software reuse repository[J].International Journal of Computer Applications,2011,27(3):33-36.
[5]GONG An,LIU Huashan,NIU Qiuli,et al.Analysis of clusteringcomponents based on XML documents similarity[J].ComputerEngineering&Design,2009,30(2):507-510(in Chinese).[龚安,刘华山,牛秋丽,等.基于XML文档相似性的构件聚类分析[J].计算机工程与设计,2009,30(2):507-510.]
[6]CHEN Dehua,HAN Zhongming,LE Jiajin.Research on clus-tering software components based on similarity analysis[J].Journal of Chinese Computer Systems,2005,26(12):2207-2210(in Chinese).[陈德华,韩忠明,乐嘉锦.基于相似性分析的软件构件聚类研究[J].小型微型计算机系统,2005,26(12):2207-2210.]
[7]Markus V D,Steffen B.Combining clustering and pattern de-tection for the reengineering of component-based software sys-tems[J].ACM SIGSOFT,2011(7):20-24.
[8]Camelia Serban,Andreea Vescan,Horia F Pop.A new compo-nent selection algorithm based on metrics and fuzzy clusteringanalysis[C].Berlin:Springer,2009:621-628.
[9]Sadaaki myamoto.Algorithms for fuzzy clustering[M].Ber-lin:Springer,2008:123-130.
[10]Dumais S T.Using latent semantic analysis to improve infor-mation retrieval[C].Proceedings of the ACM Conference onHuman Factors in Computing Systems.New York:ACMPress,1988:281-285.
[11]Shanghai component library[DB/OL].[2011-04-05].ht-tp://www.sstc.org.cn/(in Chinese).[上海构件库[DB/OL].[2011-04-05].http://www.sstc.org.cn/.]
[2]REN Yaopeng,CHEN Lichao,ZHANG Yingjun,et al.Im-proved method of component clustering based on latent semanticanalysis[J].Computer Engineering,2011,37(4):67-69(inChinese).[任姚鹏,陈立潮,张英俊,等.基于潜在语义分析的构件聚类改进方法[J].计算机工程,2011,37(4):67-69.]
[3]ZHANG Yingjun,REN Yaopeng,CHEN Lichao,et al.Com-ponent clustering algorithm based on semantic similarity and op-timization[J].Computer Engineering&Design,2010,31(11):2531-2535(in Chinese).[张英俊,任姚鹏,陈立潮,等.基于语义相似度与优化的构件聚类算法[J].计算机工程与设计,2010,31(11):2531-2535.]
[4]Kamalraj R,Rajiv Kannan A,Ranjani P.Stability-based com-ponent clustering for designing software reuse repository[J].International Journal of Computer Applications,2011,27(3):33-36.
[5]GONG An,LIU Huashan,NIU Qiuli,et al.Analysis of clusteringcomponents based on XML documents similarity[J].ComputerEngineering&Design,2009,30(2):507-510(in Chinese).[龚安,刘华山,牛秋丽,等.基于XML文档相似性的构件聚类分析[J].计算机工程与设计,2009,30(2):507-510.]
[6]CHEN Dehua,HAN Zhongming,LE Jiajin.Research on clus-tering software components based on similarity analysis[J].Journal of Chinese Computer Systems,2005,26(12):2207-2210(in Chinese).[陈德华,韩忠明,乐嘉锦.基于相似性分析的软件构件聚类研究[J].小型微型计算机系统,2005,26(12):2207-2210.]
[7]Markus V D,Steffen B.Combining clustering and pattern de-tection for the reengineering of component-based software sys-tems[J].ACM SIGSOFT,2011(7):20-24.
[8]Camelia Serban,Andreea Vescan,Horia F Pop.A new compo-nent selection algorithm based on metrics and fuzzy clusteringanalysis[C].Berlin:Springer,2009:621-628.
[9]Sadaaki myamoto.Algorithms for fuzzy clustering[M].Ber-lin:Springer,2008:123-130.
[10]Dumais S T.Using latent semantic analysis to improve infor-mation retrieval[C].Proceedings of the ACM Conference onHuman Factors in Computing Systems.New York:ACMPress,1988:281-285.
[11]Shanghai component library[DB/OL].[2011-04-05].ht-tp://www.sstc.org.cn/(in Chinese).[上海构件库[DB/OL].[2011-04-05].http://www.sstc.org.cn/.]