Abstract
I worked with Dr. Caitlin Kelleher and her research team to improve the way programmers understand unfamiliar codebases. The identified solution was to implement an IDE extension called code stories, which is a tool that collects all the information from the time a codebase is created as a resource for future programmers who work on that codebase. Before creating the code stories framework, data on YouTube coding livestreams were collected and analyzed in order to determine the pertinent information to include in code stories. To do this, I noted all the web activities, coding activities, and goals of a professional programmer working through a coding problem on a livestream, and stored the information in a SQLite database. I used the database to create a heat map that identified the number of edits undergone by each line in a programmer’s code. This heat map was used to qualitatively analyze the data by identifying problem and success areas in the programmer’s code, and by comparing heat maps between programmers.
Partner and Advisor
Partner
Dr. Caitlin Kelleher, Associate Professor and Researcher in the Department of Computer Science and Engineering at Washington University in St. Louis
Advisor
Dr. Dorothy Wang, Lecturer in the Department of Electrical & Systems Engineering at Washington University in St. Louis