The Impact of Generative AI on Code Expertise Models: An Exploratory Study

This repository contains the artifacts for reproducing the results of the study entitled "The Impact of Generative AI on Code Expertise Models: An Exploratory Study" accepted at the "Brazilian Symposium on Software Engineering - 2025". The full text of the accepted paper is available at the provided link. This repository includes the following files: chatgpt_shared_links_code.json — JSON file containing ChatGPT shared links found in GitHub source code files, with both the generated code from ChatGPT and the integrated code found in the project. truck_factor_repositories.csv — CSV file with metadata about the repositories and commits analyzed for the Truck Factor simulations. impact_analysis.zip — A compressed R project with data and analysis scripts: doe_impact_script.R: Script to simulate the impact of GenAI usage on developer expertise values (DOE). truck_factor_impact_script.R: Script to evaluate the effect of GenAI-generated code on Truck Factor values. truck_factor_values.csv: Contains the original Truck Factor values for each software project. This serves as the baseline for comparison in the GenAI impact analysis. truck_factor.csv: Contains the Truck Factor values computed under different Generative AI (GenAI) usage scenarios. This file is used to assess how varying levels of GenAI impact the concentration of knowledge in source code. contributor_file_doe_no_effect.csv: Contains Developer Expertise (DOE) values calculated without considering the influence of GenAI. Represents the original distribution of expertise based purely on developer contributions. contributor_file_doe.csv: Contains Developer Expertise (DOE) values adjusted to reflect the impact of GenAI usage. The adjustment is based on a uniform code copy rate using the average percentage of GenAI-generated code. Requirements: These scripts were tested using: R version: 4.5.1 All necessary libraries will be installed automatically within each script using standard install.packages() calls. Hardware requirements: Standard modern desktop with at least 4 GB RAM and 500 MB free disk space. No special hardware or non-standard peripherals required. Instructions for Running the R Data Analysis Scripts on Linux: Install R Follow the installation instructions for your Ubuntu version at: https://cran.r-project.org/bin/linux/ubuntu/fullREADME.html Download and Extract the Analysis Package Download the impact_analysis.zip file. Extract it and open a terminal inside the extracted impact_analysis folder. Run the Analysis Scripts From within the impact_analysis folder, execute the following commands: Rscript truck_factor_impact_script.R Rscript doe_impact_script.R View the Results The output files will be generated in the results folder.