Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Learn about the critical issue of stale reviewer recommendations in code review systems through this 17-minute conference talk that examines how inactive reviewers continue to be suggested by automated tools. Discover research findings on the prevalence of stale recommendations across five different reviewer recommendation tools (LearnRec, RetentionRec, cHRev, Sofia, WLRRec) tested on three major open-source projects, revealing that 12.59% of incorrect recommendations involve reviewers who are no longer active. Explore how top reviewers often dominate these stale recommendations, with the top-3 reviewers accounting for half of stale cases in 15.31% of instances, and understand the extreme cases where some reviewers are suggested up to 7.7 years after they've left the project. Examine the proposed filtering strategy based on recent activity that can reduce staleness by 21.44%–92.39%, while considering the trade-off of potentially shifting more review load to currently active reviewers, providing valuable insights for improving code review processes and reviewer recommendation systems.
