USING LATENT SEMANTIC INDEXING TO IMPROVE COMPONENT IDENTIFICATION
An increasing number of software systems is developed using component technologies such as COM, CORBA,or EJB.Since the introduction of component-based development (CBD), an effective component identification technique is an important factor for successful CBD projects.Obtaining components from source code can help designers and developers to understand and optimize the software system.Currently there exist many approaches to identify component,which adopt different component metrics.However, these approaches are not satisfying in the precision.This paper provides approach which introduces latent semantic indexing (LSI) technology to component identification.The approach takes into account not only the component coupling,cohesion, dependency, interface, granularity, and architecture but also the semantics of the code.We also provide a case study, in which the proposed method improves the reasonableness of the extraction results.
Components Component Metrics Reverse Engineering Latent Semantic Indezing
Feng Liu Chao Liu
Software Engineering Institute,Beihang University 100191,P.R.China
国际会议
2009 International Symposium on Computer Science and Technology(2009 中国宁波国际计算机科学与技术学术大会)
宁波
英文
59-63
2009-11-20（万方平台首次上网日期，不代表论文的发表时间）