基于源码模式挖掘的软件辅助开发技术研究

Based on the Source Code Pattern Mining Software Assist Development Technology Research

由于软件代码的复杂性,对于不了解框架的新手,很难利用开源社区中的代码来开发软件。因此,利用数据挖掘技术挖掘现有代码中的编程模式成为研究热点。文中介绍了频繁项挖掘Apriori算法,并提出了基于源码模式的软件辅助开发方法。它能够根据用户输入的关键字来智能匹配类库中的特定父类,挖掘基于此父类的编程模式,给出优先要重写的方法以及关联规则。实验结果表明,新手可以利用这些编码建议,快速学习一个新的框架,提高开发效率。

With the explosive growth of the open source community software code,software reuse is being more and more attention. Because of the complexity of the software code. To the novice that do not understand the framework,it is difficult to use the code from the open source community to develop software. As a result,using data mining technology to dig programming patterns from the existing code become a research hotspot. This paper mainly introduces the frequent items mining Apriori algorithm,and put forward the based on the source code pattern mining software assist development method. It can be based on user input keywords to intelligent matching to the super class,mining programming patterns based on the super class,give the methods that priority to override as well as the association rules. The experimental results show that the novice can use these coding suggestions,fast learning a new framework and improve the development efficiency.