摘要
软构件的检索方法一直是构件库研究的热点,而构件的合理分类是实现构件高效检索的基础。常用的基于刻面分类的构件描述方法,由于受到主观因素的影响,难以实现构件高效检索。针对这一问题,本文通过利用潜在语义分析模型,提出了一种基于潜在语义分析的构件聚类方法,在一定程度上降低了刻面分类的主观性因素,有效地提高了构件检索的效率和准确性。与采用向量空间模型的构件聚类相比较,实验表明该算法具有较好的构件聚类效果。
Software component retrieving has been a hot spot for the research of component library, while the reasonable classification of component is the foundation of components＇ efficient retrieving. Due to subjective factors, faceted classification description in common use is difficult to achieve efficient retrieval. To solve this problem, this paper proposes a component clustering method based on the model of latent semantic analysis, which to some extent reduces the subjectivity of faceted classification, and improves the efficiency and accuracy of component search. And compared with the use of vector space model, the experiments show that the model of latent semantic analysis can achieve better clustering effect.
出处
《微计算机信息》
2010年第21期166-168,共3页
Control & Automation
基金
山西省自然科学基金资助项目
基金申请人:陈立潮
任姚鹏
张英俊等
项目名称:基于语义的构件检索关键技术研究
基金颁发部门:山西省科技厅(2009011022-1)
关键词
构件
刻面分类描述
潜在语义分析
向量空间模型
K-均值聚类
component
faceted classification description
latent semantic analysis （LSA）
vector space model （VSM）
k-means clustering
