Reproduce “OSS License Identification at Scale: A Comprehensive Dataset Using World of Code” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Mahmoud Jahanshahi, David Reid, Adam McDaniel, Audris Mockus
Researchers at the University of Tennessee, Knoxville, created a comprehensive dataset for open-source software license identification, leveraging the World of Code infrastructure to map over 5 million distinct license texts to projects with commit timestamps. This resource achieved a verified accuracy of 92.08% and enables dynamic analysis of license adoption and compliance across the software supply chain.
Key insights
- A combined approach using large-scale code indexing (WoC) and a robust text fingerprinting algorithm (Winnowing) can effectively identify OSS licenses at scale, accounting for text variations.
- The 85% similarity threshold for Winnowing signatures reliably distinguishes true license matches from non-license content or unreliable partial matches, ensuring high data quality.
- Integrating commit timestamps into project-to-license mappings provides a crucial temporal dimension for dynamically analyzing license adoption, evolution, and compliance throughout the software supply chain.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.