DISO: Defence, Intelligence and Security Ontologies

The DISO dataset aims to inform the UK Government intelligence community about the public ontologies relevant to defence and national security and related areas. The ontologies in DISO were identified and obtained during an ontology search exercise undertaken in November and December of 2025. Each ontology is accompanied by documentation that provides information such as a full name, a short description, relevant context, a web presence, and the source from which the ontology was obtained.

Most of the ontologies in DISO cluster into categories (or subdomains) of the umbrella domains of defence, intelligence and security. The structure of this repository reflects this clustering. Sometimes, an ontology fits into more than one cluster (subdomain). However, each ontology appears only once in the repository: in the cluster where the fit is deemed strongest.

The domains of defence, intelligence and security have been interpreted broadly. The connection between some of the subdomain clusters and the umbrella domains may appear tenuous. In the README files for these subdomains, we provide information and cite sources to explain and justify our decision to include the subdomain within DISO.

Some of the ontologies in DISO are distributed as networks of component ontologies, woven together with owl:imports statements. For some of these, we have merged the component ontology files into a single ontology file, using the Protege ontology editor.

The diso-compact directory bundles a canonical file per ontology. If an ontology belongs to multiple clusters, a copy is placed in each cluster to which it belongs. As such, the compact distribution contains some duplicate ontologies. The compact distribution is packaged for the convenience of downstream consumers, such as the DISO ontology matching pipeline.

This repository serves as the resource layer for a small ecosystem of repositories that support an OAEI track for defence, intelligence, and security ontologies.

DISO-mappings is a Python pipeline that consumes the DISO compact distribution and produces pairwise alignments between selected ontology pairs. It uses a configurable set of ontology matching systems (AML, LogMap, LogMapLt, BERTMap, and BERTMapLt by default; with an extendable Matcher base class that enables custom external matchers). The pipeline aggregates the per-system alignments into an unverified silver-standard consensus alignment via a family-based voting mechanism. This alignment is then manually verified, adapted and used as the basis for a partial reference alignment. To reproduce our alignments, follow the steps within the repository’s make workflow.

DISO-oaei is the home page for the OAEI track and will describe the evaluation process involving above-mentioned reference alignment.