Abstract: Feature location is one of the prerequisites for the sooth realization of software evolution intent. There are some problems in current location method, such as low precision, huge location space and low automatic degree. For these problems, this paper proposes a feature location method based on behavior knowledge and topic knowledge. This method combines the behavior knowledge with topic knowledge to reduce the state space of feature location, and then realizes the feature location by analyzing the similarity between topic knowledge and software function attributes. The experimental results show that compared with traditional textual feature location method, the proposed method has higher recall, precision and comprehensive performance. Finally, the experiments validate the effectiveness of the proposed method.

Key words: feature location, software evolution, behavior knowledge, topic knowledge

摘要： 软件特征定位是成功实施软件演化意图的前提条件。针对当前特征定位方法准确率低，定位空间庞大和自动化程度不高等问题，提出了一种面向行为主题的混合特征定位方法。该方法将软件系统的行为知识和主题知识进行有机整合，实现了行为知识对主题知识的约简，并最终通过分析主题知识与软件功能属性间的相似性实现特征定位。实验结果表明，与传统的基于文本的特征定位方法相比，该方法具有较好的查全率、查准率和综合性能，从而验证了该方法的有效性。

关键词: 特征定位, 软件演化, 行为知识, 主题知识