I am a master’s student in Cheriton School of Computer Science at the University of Waterloo and I am a member of the SWAG Lab and the Software REBELs Lab. I am supervised by Dr. Michael Godfrey and Dr. Shane McIntosh. My research interests includes mining software repositories, software defect prediction, and code review analysis. My research area is to improve the situational awareness for modern code review. Before starting my master’s program, I received my bachelor’s degree (BCS) in Cheriton School of Computer Science at the University of Waterloo.
Title: Studying the Impact of Risk Assessment Analytics on Risk Awareness and Code Review Performance
Principal Investigator: Shane McIntosh
Co-Principal Investigator: Michael Godfrey
Student Investigator: Xueyao (Eve) Yu
Co-Investigator: Filipe Cogo
Description: Code review is broadly considered an essential step in software quality assurance. However, defects often slip through the review process undetected. Prior work suggests that a lack of awareness of the defect proneness of changes is the main associated factor with defects that slip through code review. Indeed, it is difficult for even the most seasoned developers to retain an awareness of the riskiness of code changes. Changes for which riskiness has been underestimated are likely to be insufficiently reviewed, which in turn may allow defects to slip into the codebase. In this study, we implement a risk assessment tool based on the notion of historical risk analysis and we seek to examine the extent to which risk assessment analytics impact developer awareness of the riskiness of code changes, as well as reviewer effectiveness and efficiency in defect detection during code review.
Participant Recruiment: We plan to perform our study on the Qt codebase, which has a high rate of code review coverage and has been widely used as a target system in prior related studies [1, 2, 3, 4]. We are currently recruiting developers from Qt Base to participate a short user experiment (30-60 minutes). The experiment will be performed online and asynchronously using a simple web application (hosted on Heroku). If the experiment goes well, we'll release our risk assessment tool as a (free) plug-in to Gerrit.
If you are interested in participating this study, please sign up here.