Sharif Ahmed – Computer Science
CCP 368 or Zoom (passcode: 050178)
Title: An analysis on useful comments in modern code review process
The Code Review process is advancing several fields of Computer Science by enabling high-quality and reliable software, ensuring the reliability and safety of AI systems, and uncovering security vulnerabilities in cyber systems. In collaborative software development, the peer code review process is beneficial only if the reviewers provide useful comments and there are automatic tools to help reviewers with suggestions to offer useful comments. Researchers have found code review comments to be a paramount component of the Modern Code Review process in the industry and open-source development. Therefore, it is important to ensure that code review comments serve their purposes. This paper investigates how existing research defines the usefulness of code review comments, mines and annotates datasets, analyzes factors from different aspects, and uses machine learning classifiers to automatically predict the usefulness of code review comments. It then discusses the open problems and challenges in recognizing useful code review comments.
Committee: Nasir Eisty, Ph.D., (Chair), Elena Sherman, Ph.D., Francesca Spezzano, Ph.D., Jyh-haw Yeh, Ph.D (CompEE)