A Systematic Mapping Study on Software Comments Analysis

Abstract

Mining software repositories has been used as an important tool to support software engineering research. Recent studies indicate that code comments are one of the most explored objects of analysis in the area. Objective: This work investigates how analysis of comments has been used to support software engineering activities by identifying its purposes, focuses, techniques, tools, evaluation methods, and the research type performed in the area. Method: We performed a systematic mapping study of the literature that considered papers from 1990 to 2016. Results: We analyzed 36 primary studies. The collected data pointed out that comment analysis has been used mainly for understanding and identifying the quality of software artifacts. The Dictionary/Vocabulary and Natural Language Processing are among the most used techniques, and most of them are performed in a semiautomatic way. We also organized a set of tools that have been used for mining software comments. Most of the primary studies are a solution proposal paper. Regarding evaluation methods, we found that experiments and case studies are the most considered. Conclusion: The results of this mapping study can help to identify points that still require further investigation in comment analysis research.