An Automated Approach to Assessing Code Quality via Bug Reports

Abstract

Mining Software Repositories (MSR) can serve various purposes such as analyzing code coverage, predicting code changes. One of software repositories is bug repositories. However, bug repositories, i.e. bug reports, has not been made full use of due to that many researchers or practitioners lay little emphasis on mining bug reports. It is because they may not be aware of its value or practical meaning. In order to take advantage of bug reports, in the paper, an original approach has been put forward to assessing code quality or at least estimating the status of software by mining bug reports, which is automated. The results of experiments show the effectiveness of this novel approach. To author’s best knowledge, it was the first time to make such full use of mining bug reports in an automated way.