Semantic extraction pipeline

Semantic extraction pipeline is a Python pipeline he built during his joint research at JAIST in 2025. It extracts structured semantic facts for safety testing of automated driving systems, alongside the ADS safety-testing ontology.

How it ran

• Validation: the extracted facts were checked with SPARQL queries.
• Scale: the work ran on JAIST's KAGAYAKI HPC and PBS job orchestration cluster as batch jobs submitted through PBS, the cluster's job scheduler.

The PBS commands he used there (qsub, qstat, qdel) later turned up as shortcuts in his command-line tool mops.

See also

• text2kgraph, his earlier text-to-graph prototype
• Ontology and knowledge graphs