Alaa Ismail, a software engineering student, presented her research paper, "Analyzing Reading Behavior across Source Code and Stack Overflow for Method and Class Summarization," at the Eye Movements in Programming Workshop, co-located with the ACM Symposium on Eye Tracking Research and Applications (ETRA), held in Marrakech, Morocco, in June 2026.
The paper was co-authored with Zachary Kozak and Dr. Bonita Sharif. As first author, Ismail led the research effort and manuscript preparation. The study reexamines an existing eye-tracking dataset focused on Java method and class summarization, providing new insights into how developers read and process information from both Java source code and Stack Overflow.
As part of the analysis, Ismail generated detailed scarf plots for each task in both the Java source code and Stack Overflow contexts and conducted a qualitative examination of reading behaviors associated with both correct and incorrect responses. The study also explored the search terms developers used when seeking information on Stack Overflow.
The findings revealed that participants generally spent the most time viewing Java source code, while exhibiting longer fixation durations on Stack Overflow content. Results further suggest that developers’ gaze behavior and summarization accuracy within Stack Overflow depend on the information available for a given task, as both scan paths and accuracy scores varied across tasks. Additionally, the study found that participants who made more fixations on key areas of interest, and who transitioned more frequently between those areas, tended to achieve higher accuracy scores. This pattern was consistent across both the Java source code and Stack Overflow environments.
This research was conducted as part of Ismail's undergraduate Honors Thesis. She worked under the supervision of Dr. Sharif during fall 2025 and spring 2026.
Ismail graduated from the School of Computing's software engineering program in spring 2026 and is now employed as a full-time software engineer at a local software development firm.
