A model to detect readability improvements in incremental changes

Abstract

Identifying source code that has poor readability allows developers to focus maintenance efforts on problematic code. Therefore,the effort to develop models that can quantify the readability ofa piece of source code has been an area of interest for softwareengineering researchers for several years. However, recent researchquestions the usefulness of these readability models in practice.When applying these models to readability improvements that aremade in practice, i.e., commits, they are unable to capture theseincremental improvements, despite a clear perceived improvementby the developers. This results in a discrepancy between the modelswe have built to measure readability, and the actual perception ofreadability in practice.In this work, we propose a model that is able to detect incremental readability improvements made by developers in practice withan average precision of 79.2% and an average recall of 67% on anunseen test set. We then investigate the metrics that our modelassociates with developer perceived readability improvements aswell as non-readability changes. Finally, we compare our modelto existing state-of-the-art readability models, which our modeloutperforms by at least 23% in terms of precision and 42% in termsof recall.