All data concerning my thesis

Title: Code and Supplementary Files for "Leveraging Large Language Models as Requirements Elicitation Interview Bots"

Description:
This repository contains all code, supplementary plots, and select data files used in the master’s thesis, "Leveraging Large Language Models (LLMs) as Requirements Elicitation Interview Bots." The study explores the customization and application of LLMs, specifically GPT-4o and Llama 3.1, in automating interview processes in Requirements Engineering (RE).

Contents:

• Python Scripts: Scripts used for data preprocessing, analysis, and generation of plots. These scripts support the findings, allowing for replication of bot customization, review analysis, and performance comparisons across various prompt engineering approaches.
• Supplementary Plots and Visuals: Visualizations supporting the thesis findings, including mistake type distributions, error rate comparisons across LLM versions, recall/precision analysis, and boxplots for requirements elicitation accuracy.
• Datasets: While select datasets are included, the dataset used for fine-tuning is not provided, as it includes interview data sourced from Ferrari et al.’s study, "Learning Requirements Elicitation Interviews with Role-playing, Self-assessment and Peer-review," and is not shareable under current permissions.
• ReadMe File: A guide to understanding and running the provided code, detailing dependencies, setup instructions, and folder structure for reproducibility.

These materials support the analyses presented in the thesis and facilitate further exploration of LLM customization for RE interview tasks.