摘要
演化性是软件的固有属性。整个软件演化过程是由一系列活动构成,同时该过程具有复杂性、持续变化性、自治性、并发性、迭代性等特点。这些属性决定了软件演化过程是一个动态的、难以度量的过程。基于markov链理论,采用状态转移矩阵和权重因子等技术构造了关于软件演化过程对资源开销的度量模型。软件演化过程度量体系的建立对进行软件演化管理、控制、评估具有重要意义,同时更为重要的是该体系的建立为软件演化的自适应演化奠定了基础。
Evolution is an intrinsic property of software, and the process of software evolution consists of a series of activities. Meanwhile the evolution involves many features, includes complexity, continue change, self regulation, iteration, concurrence, etc. These inherent features make the software evolution process dynamic and hard to evaluate. Based on Markov chain theory, in this paper it proposed an evaluation model of software evolution process for resource cost using State Transform Matrix （STM） and weight factor technology. All that we did are of credible foundation for managing, controlling and evaluating software evolution, Moreover, the establishment of the model lays a foundation for adaptive evolution of software evolution based on matrix shift in computer.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第9期17-18,92,共3页
Computer Applications and Software
基金
国家自然科学基金资助项目(60463002)
关键词
软件演化
马尔可夫链
生产率
状态转移矩阵
权重因子
Software evolution Markov chain Productivity State transform matrix Weight factor
