基于开发过程数据的软件可维护性评估  

Software Maintainability Assessment Based on Development Process Data

作　　者：郝学良[1] 朱小冬[1] 叶飞[1] 刘丽[2] 

机构地区：[1]军械工程学院装备指挥与管理系,河北石家庄050003 [2]石家庄机械化步兵学院教研部,河北石家庄050083

出　　处：《军械工程学院学报》2016年第6期41-45,共5页Journal of Ordnance Engineering College

摘　　要：针对软件维护性影响因素多、度量难度大、软件维护实验成本高,且不可避免地产生霍桑效应等问题,提出一种基于开发过程数据挖掘的维护性评估方法,建立了基于平均功能块维护时间的维护性评估模型.通过对软件配置管理库、文档等数据源进行数据收集、整理、分析,将软件维护时间作为维护性验证参数,实现了对软件维护性的定量评估.最后,利用实际的软件工程项目验证了方法的可行性和有效性.The factors affecting software maintainability are big in number,difficult to measure,high in its experimental costs and they inevitably result in the Hawthorne Effect.To tackle these problems,an evaluation method based on development process data mining was presented and software evaluation model based on Mean Function Block Maintenance Time was built.Through software configuration repository and documentary file, maintainability data was collected,transacted and analyzed to evaluate the software maintainability of project.Finally,a real software project is used to validate the method and the project proves that this proposed method is feasible and effective.

关 键 词：软件工程 数据挖掘 维护性评估 维护时间 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]