基于模糊聚类的构件检索方法  

Component retrieval model with Fuzzy cluster analysis

作　　者：董跃华[1] 马亚飞[1] 

机构地区：[1]江西理工大学信息工程学院,江西赣州341000

出　　处：《南昌大学学报（理科版）》2012年第4期398-401,共4页Journal of Nanchang University(Natural Science)

基　　金：国家自然科学基金资助项目(71061008)

摘　　要：在构件的检索过程中,由于用户对于构件的描述形式或者机制不是很理解,因此很难把自己的需求以专业的术语或者表达形式表示出来,从而影响了检索的效率。引入了刻面权重的定义,将用户的需求有效的具象化,并提出了基于模糊聚类分析的构件检索方法,利用一定的聚类准则将构件库里的构件集合划分为不同的类别,降低构件检索的规模,提高构件检索的效率,同时具有较好的查全率和查准率。实验结果证明了该方法的可行性与有效性。Users usually couldn＇t demonstrate their requirements clearly in a technical way or an explicit expression,because of the low intelligent of components＇ description form or mechanism in the process of component retrieval.This would influence the efficiency of retrieval.This paper introduced some related concept of faceted classification schema and the definition of facet-weight to quantize users＇ requirements.We proposed a new component retrieval model with fuzzy cluster analysis.The schema improved the query efficiency by reducing the scale of component retrieval with certain clustering criteria,especially the retrieval of large-scale component-library.The experiment showed that the schema was efficient and feasible.

关 键 词：刻面 刻面权重 构件检索 构件库 软件复用 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]