摘要
该文提出了一种改进的软件项目开发风险管理模型。该模型在贝叶斯网络的建模过程中以样本数据集为基础进行结构学习和参数学习,建立更符合实际软件项目特征的贝叶斯网络。同时,进一步完善了软件项目开发风险管理流程,并利用贝叶斯网络的信念更新过程实现动态软件项目风险管理。经实践检验,该改进模型能够更有效地对软件项目开发过程中的风险进行管理,提高软件开发的成功率。
An improved risk management model in software project development is presented in this paper.In the process of Bayesian Networks (BNs) modeling,the model utilities structure learning and parameter learning based on the sample data for building the BNs in accord with the feature of software project.In addition,the flow of risk management in software project development is adjusted and dynamic risk management of software project is realized using BNs belief update.Practice proves that the improved model can effectively accomplish the work of risk management in software project development.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第21期1-3,7,共4页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(the National Natural Science Foundation of China under Grant No.70371046)。
关键词
软件项目
风险管理
贝叶斯网络
结构学习
参数学习
software project
risk management
Bayesian Networks (BNs)
structure learning
parameter learning