在软件易变性上下文中类规模对面向对象度量的影响分析  

Potentially confounding effect of class size on validity of object-oriented metrics in context of software change-proneness

作　　者：吴方君[1,2] 

机构地区：[1]江西财经大学信息管理学院,南昌330032 [2]江西财经大学江西省高校数据与知识工程重点实验室,南昌330032

出　　处：《计算机应用研究》2017年第8期2417-2421,共5页Application Research of Computers

基　　金：江西省自然科学基金资助项目(20142BAB207010);江西省教育厅科技项目(GJJ160427)

摘　　要：软件易变性预测主要通过软件的内部特性,即软件度量值来刻画、预测的,是软件工程中热点方向之一,在提高软件质量、控制软件成本方面起着非常重要的作用。虽然软件易变性预测在学术界取得了一系列的成绩,但在工业界尚未有成功应用的案例。从简单相关性分析与偏相关性分析和关联规则挖掘的角度出发甄别面向对象度量与软件易变性间相关性的真伪,明确了在软件易变性上下文中类规模对面向对象度量有潜在影响。Software change-prone prediction, one of the hot topics in software engineering domain, mainly through the internal characteristics of software, namely software metrics, to describe and forecast the external characteristics of the software, plays an important role in controlling and improving software quality and software cost balance. Although software change-prone pre- diction has made a series of achievements in the academic field, there has not yet been a successful application in the indus- trial field. This paper identified the authenticity of the correlation between object-oriented metrics and software change-prone- ness from the aspects of simple correlation analysis, partial correlation analysis and association mining. The experiment results show that there are potentially confounding effect of class size on the validity of object-oriented metrics in the context of soft- ware change-proneness.

关 键 词：软件易变性 面向对象度量 相关性分析 关联规则挖掘 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]