An Eye-Tracking Study of Programmers to Improve Automated Source Code Summarization

Paige Rodeghero, Collin McMillan, Paul "Will" McBurney, Nigel Bosch, Sidney D'Mello


Please use the resources on this page to help you verify our results and recreate our approach. An archive of all of the individual resources listed below can be found here. If you need additional help or information, feel free to contact Paige Rodeghero.

Eye-Tracking Data

Raw Eye-Tracking Data

The above compressed folder contains all of the raw eye-tracking data collected during the experiment. This data was parsed and analyzed to figure out where programmers tended to look while summarizing. Below is the compressed folder containing the Excel spreadsheets used to analyze this data.

Eye-Tracking Data Analysis

Source Code

Keyword Extraction Tool

The above compressed folder contains the source code for our new technique.

Follow-up Study Data

Follow-up SQL Dump File

The above SQL dump file contains the database for the follow-up study. The data recorded here was analyzed to determine whether the state-of-the-art tool or our own technique was better. Below are the Excel spreadsheets used to analyze this data.

Follow-up Kmin Analysis

Follow-up Overlap Analysis

"Stubborn" Keywords Data

The files below are archives containing the analysis and the scripts used for the "stubborn" keywords section.

"Stubborn" Keyword Scripts

"Stubborn" Keyword Analysis

Paige Rodeghero
Dr. Collin McMillan
Paul "Will" McBurney
Nigel Bosch
Sidney D'Mello