This modern code review has shown increasing adoption both in industrial and open source contexts [7], [8], [9] For example, in January of 2015, more than 50,000 Microsoft employees practiced tool-based code review. A recent survey [10] found that on average developers spend about six hours per week preparing code for review or reviewing others’ code. Therefore increasing the effectiveness of code review practices is beneficial for ensuring developers’ time (both the author
and the reviewer) is spent wisely. The main building blocks of code reviews are comments that reviewers add as feedback and suggestions for change that the code review author can address. The usefulness of those comments highly influence the effectiveness of the code review practices. Comments that point out bugs, suggest improved ways of solving problems, or point out violations of team practices can help the author submit a higher quality change to the codebase, improve the author’s development skills, or both. Other comments may contain incorrect information or may ask questions that are not relevant and require the author’s time to respond to without improving the code.