More Accurate Recommendations for Method-Level Changes
This addresses the time-consuming and error-prone task of applying similar code changes in software development, offering an incremental improvement over existing tools.
The paper tackles the problem of low accuracy in code change recommendations due to code movements, presenting ARES which achieves an average accuracy of 96% on developer-performed changes while maintaining competitive precision and recall.
During the life span of large software projects, developers often apply the same code changes to different code locations in slight variations. Since the application of these changes to all locations is time-consuming and error-prone, tools exist that learn change patterns from input examples, search for possible pattern applications, and generate corresponding recommendations. In many cases, the generated recommendations are syntactically or semantically wrong due to code movements in the input examples. Thus, they are of low accuracy and developers cannot directly copy them into their projects without adjustments. We present the Accurate REcommendation System (ARES) that achieves a higher accuracy than other tools because its algorithms take care of code movements when creating patterns and recommendations. On average, the recommendations by ARES have an accuracy of 96% with respect to code changes that developers have manually performed in commits of source code archives. At the same time ARES achieves precision and recall values that are on par with other tools.