Insurance Data Ontology and Knowledge Graphs: Identifier Governance for Insurers and Reinsurers
🔒 This course requires registration
To access this course and all our learning materials, please register for the AI Fluency programme.
Insurance entity data is an identifier problem before it is anything else. This course teaches you to build and govern an insurance ontology and knowledge graph for insurer and reinsurer entity data, using the same method we applied to every authorised insurer in Europe.
It is grounded in a real open study rather than a worked example. We took the complete EIOPA Register of Insurance Undertakings, 33,924 rows covering every EU and EEA authorised insurance and reinsurance undertaking, joined it to the global GLEIF identifier system, and cross-checked it against the German national register. That study found that 643 of 3,304 active insurers carry no Legal Entity Identifier at all, that four filed identifiers are arithmetically impossible, that 42 name entities the global register says no longer exist, and that one identifier is filed for two distinct SCOR reinsurance companies. Every technique in this course is one we used to find those defects.
What you will learn
- Model insurer and reinsurer entities, authorisations and cross-border passports in OWL 2, so lifecycle rules can actually be written
- Build SKOS registries that declare identifier scope as data, which is what stops branch structures raising false duplicates
- Write SHACL shapes that turn data policy into executable constraints, graded by severity so the output is a work queue rather than noise
- Validate Legal Entity Identifiers with ISO 7064 check digits, and understand why truncation and transposition need different fixes
- Reconcile a supervisory register against GLEIF and against national regulators, and record disagreement instead of silently resolving it
- Resolve treaty and facultative counterparties reliably, including the entity collapse failures that corrupt group exposure roll-up
- Lay foundations for Solvency II reporting, data lineage and quality that survive audit
Who this is for
Insurance data leaders, data architects, actuaries, and regulatory reporting teams at insurers, reinsurers, brokers and supervisors. No prior semantic web experience is assumed; the ontology concepts are taught from insurance problems rather than from theory.
The open artefacts behind the course
Everything the course is based on is public and reproducible. Read the full study, An open ontology for insurance and reinsurance, tested against every insurer in Europe, or clone the ontology, SKOS registries, SHACL suite and pipeline from github.com/fabio-rovai/insurance-register-ontology. Ontology and documentation are CC BY 4.0, code is MIT, and no licensed data is used anywhere.
If you want this method applied to your own entity data, write to fabio@thetesseractacademy.com.