These past two weeks have been extremely hectic, because of YSU's last week of classes and finals. I worked on interpreting Alyssa's results and writing our final report. We found that novice developers assign less tags than expert developers. They as spend more time on text and titles as task complexity increases to assign tags. Expert developers tend to not only assign more tags, but they also spend more time on code when task complexity increases. They also more accurately assign tags than novice developers. These trends are important for informing tag prediction models, because expert developers have distinct trends from novices and more accurately predict tags. Therefore, using expert eye gaze pattern features on Stack Overflow documents to predict tags may prove lucrative.
I also finished my senior capstone report and presented the work. This work was an extension of the research from last year's CREU project. We explored the same eye-tracking dataset using sequential analysis techniques to study/predict developer expertise with different measures. These techniques also put emphasis on the order in which developers navigate code. This work is a more mature version of my QUEST poster and the paper we submitted to MSR a few months back. Here is a picture from my presentation:
Finally, I graduated Summa Cum Laude with a BS in mathematics and computer science from Youngstown State University today. Here is a picture with Dr. Bonita Sharif and Dr. Alina Lazar:
- Finished the final report for CREU
- Presented my capstone project extending the eye-tracking research we performed in last year's CREU project.