点击次数：

所属单位：计算机科学与技术学院/人工智能学院/软件学院

发表刊物：J Syst Software

摘要：Code comments are crucial to program comprehension. In this paper, we propose a novel approach ContextCC to automatically generate concise comments for Java methods based on neural networks, leveraging techniques of program analysis and natural language processing. Firstly, ContextCC employs program analysis techniques, especially abstract syntax tree parsing, to extract context information including methods and their dependency. Secondly, it filters code and comments out of the context information to build up a high-quality data set based on a set of pre-defined templates and rules. Finally, ContextCC trains a code comment generation model based on recurrent neural networks. Experiments are conducted on Java projects crawled from GitHub. We show empirically that the performance of ContextCC is superior to state-of-the-art baseline methods. © 2019 Elsevier Inc.

ISSN号：0164-1212

是否译文：否

发表时间：2019-10-01

合写作者：Yan, Xin,邢岩,杨文华,Chen, Taolue,黄志球

通讯作者：周宇