An Interactive Tool to Improve Program Readability for Novice Students

In recent years, higher education institutions have increasingly adopted practical software development, such as project-based learning (PBL) and pair programming, in their classes. However, it is difficult for novice students to consider the readability of the source code because just converting the requirement specification into a program is challenging for them. Metrics based on machine learning of human evaluations have been proposed to evaluate the readability of source code quantitatively. This paper presents an interactive support tool that displays readability scores for source code written by novice programmers using one of those metrics by Buse and Weimer (B&W). Using this tool, we expect to make novices aware of things that professional developers naturally do, such as inserting comments and blank lines. A brief evaluation experiment with five fourth-year university students confirmed the effectiveness of the proposed tool. In addition, we discuss issues that may arise when this tool is used for educational purposes.