A Methodological Framework for LLM-Based Mining of Software Repositories

Vincenzo De Martino, Joel Castaño, Fabio Palomba, Xavier Franch, Silverio Martínez-Fernández · 2025

Read A Methodological Framework for LLM-Based Mining of Software Repositories by Vincenzo De Martino, Joel Castaño, Fabio Palomba, Xavier Franch, Silverio Martínez-Fernández (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-02233.txt.

Introduction

Vincenzo De Martino and co-authors map in 2025 how researchers use large language models to mine software repositories. Mining code archives can reveal how software is built and fixed, and AI models now automate parts of that work. But studies vary widely in method, and their rigor is hard to judge.

The team mixed a rapid review of papers with a survey of researchers in the field. They found fifteen method patterns, nine main threats to sound results, and twenty-five ways to counter those threats. From this they built a framework of six stages and twenty-three steps, each tied to its risks and fixes. The authors hope it makes this young field more open and repeatable.

Catalog record

CREATORVincenzo De Martino, Joel Castaño, Fabio Palomba, Xavier Franch, Silverio Martínez-Fernández
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE166 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-02233

Browse the full library · About the Tower of Babel