The SV-COMP Evaluation Scripts project is a comprehensive toolkit designed to parse, aggregate, and visualize benchmark datasets from the prestigious International Competition on Software Verification (SV-COMP).
Capabilities & Overview
- Data Ingestion: Parses large-scale XML/CSV verification benchmark outputs generated by competing software verifiers.
- Scoring & Quantile Analysis: Computes competition scores, penalty deductions, and CPU execution time distributions.
- Interactive Visualization: Web frontend allowing researchers to compare verifier performances across standard categories and property benchmarks.