摘要
从系统源码中挖掘设计模式对软件的可理解性和可维护性具有重要意义。基于模糊理论,提出一种模式匹配方法,实现设计模式挖掘。其中,使用基于类关系的素数矩阵模型对设计模式结构及源码信息进行描述,并作为匹配的模型基础;采用聚类方法对源码模型进行优化,提高匹配效率;将模糊方法与设计模式匹配策略相结合,引入静态和动态信息,提高匹配的正确性。实验结果证明此方法在精确性和完整性方面得到了很大的提高,并且避免了对特殊模式的失效性。
Mining design patterns from source code is a very important technology for improving the intelligibility and maintainability of software.In this paper,a new matching method based on fuzzy is presented to mine design patterns.A matrix model is used to describe design patterns and source code as the basis of matching.Clustering method is adopted to optimize the source code model in order to improve the efficiency of matching.Combining fuzzy theory to patterns matching,this method introduces the static and dynamic information to enhance the accuracy of matching.Experimental results demonstrate the improvement of matching in accuracy and integrality,avoiding the invalidation for some special design patterns.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第2期150-153,共4页
Computer Engineering and Applications
关键词
设计模式
模糊匹配
模式挖掘
矩阵模型
聚类
design pattern
fuzzy matching
pattern mining
matrix model
clustering
