pygraft

Why we included this project

PyGraft is a Python library that generates synthetic schemas and knowledge graphs from a YAML configuration file, giving you realistic stand-ins when real datasets are private, scarce, or too sensitive to share. The output uses RDFS and OWL constructs and is checked for logical consistency with the HermiT reasoner, so the data behaves like genuine graph data rather than random noise. That makes it useful for benchmarking graph-based models, testing retrieval pipelines, and building evaluation datasets in domains where the original data can't be published. Researchers and practitioners who need reproducible, tunable graph data for experiments will find the configuration-driven workflow easy to drop into their existing tooling.