Archived version of ComProScanner benchmarks metadata in various models, indicating advancements in material extraction.
Archived version (v2026.02.02) of ComProScanner, which is referenced in the Digital Discovery paper. This release includes: the snapshot of the ComProScanner package, which has been referenced in the Digital Discovery paper examples folder with: minimal and test (used script for evaluation) scripts to run ComProScanner 5 years of piezoelectric materials-related journal articles' metadata collected full-text articles where d33 were mentioned as CSV and vector-database entry data related to 100 randomly chosen DOIs from the 3917 d33-mentioned articles to benchmark ComProScanner across 10 different cost-efficient LLMs all model logs and outputs for 100 test articles across 10 LLMs data related to the comparison with similar existing frameworks (Eunomia and the extraction agent by CMEG-IITR) scripts to regenerate graphs, table information and other Materials Project (MP)-related data for the paper. Latest version (v0.1.5) of ComProScanner The following link compares the changes between the archived version and the latest version: https://github.com/slimeslab/ComProScanner/compare/v2026.02.02...v0.1.5
No takes yet. Share an insight, caveat, or question.
Roy et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: