Information extraction from Italian medical reports: An ontology-driven approach.

•We present an unsupervised, ontology-driven approach to extract information from medical reports written in Italian.•The system extracts events (e.g. diagnostics tests) and their attributes (e.g. test results) without the need for annotated data.•The system is evaluated against a structured hospital database and performs well on the considered domain.•We assessed the multilingual extendibility of the proposed approach with promising results.