We align three upper ontologies that sit under UK and NATO defence data infrastructure: the Information Exchange Standard (IES), the Higher Quality Data Model (HQDM) that underpins the National Digital Twin, and Basic Formal Ontology (BFO). No public alignment between IES and HQDM existed. Promoting a hand-curated 17-correspondence crosswalk to OWL and reasoning over the complete merged ontologies with HermiT produces three results that we believe matter beyond this pair.
Figures & tables
Ontology
triples
named classes
object props
disjointness
unsat (alone)
IES ( ies-common.ttl )
4,039
511
162
0
0
HQDM ( hqdm.owl )
3,127
229
74
14
39
BFO 2020 ( bfo.owl )
1,221
35
0
19
0
Table 1: The three ontologies as measured on the vendored files. Named classes counts rdf:type owl:Class subjects with an IRI. Disjointness counts owl:disjointWith triples; none of the three uses owl:AllDisjointClasses or owl:disjointUnionOf . Unsatisfiable classes are computed by HermiT on each ontology alone.
Criterion
Result
Consistency (a model exists)
passes
New unsatisfiable classes (beyond the native 39)
100
New source-internal subsumptions (S2 violations)
218
Table 2: Naive equivalence promotion of the expert crosswalk. The consistency check, which is what deployment pipelines typically run, passes.
Source ( ies: )
Target ( hqdm: )
Verdict
S2 if reversed
Element
spatio_temporal_extent
full ≡
0
Thing
thing
full ≡
0
Event
activity
full ≡
0
PossibleWorld
possible_world
full ≡
0
ClassOfElement
class_of_spatio_temporal_extent
full ≡
0
Entity
individual
forward only
104
Table 3: The certified bridge. Prefixes are omitted: every source is ies: and every target hqdm: . The last column is the number of S2 violations the reverse edge would have introduced, which is why the row is one-directional. Every verdict is a reasoner decision, not a curator’s judgement.
Mapping for ies:Entity
Rest of bridge
IES classes collapsed
≡ bfo:material entity
absent
0
⊑ bfo:material entity
absent
0
≡ bfo:material entity
present
101
⊑ bfo:material entity
present
100
⊑ bfo:history
present
0
Table 4: The cost of one axiom. Every row merges the complete IES and BFO with ies:Element ⊑ bfo:occurrent , ies:Event ⊑ bfo:process and ies:State ⊑ bfo:occurrent held fixed, and varies only how ies:Entity is mapped. Counts are IES classes made unsatisfiable.
System (pair)
mappings
backbone
false friend
unsat targets
new unsat
S2 viol.
LogMap (IES–HQDM)
24
3/1/6
yes
2
87
103
LogMapLt (IES–HQDM)
36
3/1/6
yes
2
214
112
LogMap (IES–BFO)
2
0/1/3
—
0
0
2
certified bridge (IES–HQDM)
13
10/0/0
no
0
0
0
Table 5: Matcher outputs on the vendored pairs, evaluated with HermiT under the same criteria as the certified bridge. “Backbone” scores correct target / wrong target / missed against the pair’s gold targets: the ten expert class correspondences for IES–HQDM, the four certified bridge targets for IES–BFO. “Unsat targets” are accepted mappings whose HQDM target is one of the 39 natively unsatisfiable classes. The certified bridge row is the output of § 4.3 .
System
into HQDM (8)
into BFO (4)
total
Qwen3-Coder-30B-A3B-Instruct (8-bit)
5
1
6/12
+ IES fine-tune
6
1
7/12
Table 6: Mapping-oracle accuracy against the HermiT-certified bridge. n=12 .
The use of ontologies and knowledge graphs is becoming increasingly widespread in the defence and national security domain. Numerous ontologies have been developed through initiatives led by academia, industry, and government. Achieving interoperability across diverse defence and national security ontologies remains a major challenge due to the domain's breadth and specialisation. In this work, we analyse and document over 60 publicly available ontologies and introduce a new track for the Ontology Alignment Evaluation Initiative (OAEI). This track comprises eight matching tasks, consensus alignments and manually-curated (silver-standard) mappings. The consensus alignments are derived by aggregating the outputs of several state-of-the-art ontology alignment systems. The silver-standard is obtained from the manual validation of the consensus alignment together with a subset of the unique mappings (i.e., mappings suggested by only one system).
Jonathon Dilworth, Pedro Giesteira Cotovio, David Herron +4
City St George’s, University of London, London, UK · LASIGE, Faculdade de Ciências, Universidade de Lisboa, Portugal · Defence Science and Technology Laboratory, UK +1
In this paper we discuss the foundations behind a novel methodology for the validation of semantic mappings between different data sources based upon different foundation ontologies, where the methodology builds a framework based upon the metaphysical commitments of the ontologies. We provide as example the test case of mappings between IES and BFO, and we especially focused on providing cardinality constraints on the mappings between the two ontologies. In order to demonstrate the applicability of our method, we showcased how these principles can be operationalized through SPARQL queries validating the results of a mapping pipeline.
Giacomo De Colle, Helena Blackmore, Chris Partridge
We present Open Ontologies, an open-source ontology engineering system implemented in Rust that integrates LLM-driven construction with formal OWL reasoning and ontology alignment via the Model Context Protocol. Our primary finding is that stable 1-to-1 matching is the dominant factor in ontology alignment quality: on the OAEI Anatomy track, it achieves F1 = 0.832 (P = 0.963, R = 0.733), competitive with state-of-the-art systems and exceeding all in precision. Ablation across five weight configurations shows that signal weights are irrelevant when stable matching is applied (F1 varies by less than 0.004), while removing stable matching drops F1 to 0.728. On the Conference track, the same method achieves F1 = 0.438. On tool-augmented ontology interaction, we find a surprising result: an LLM reading a raw OWL file (F1 = 0.323) performs worse than the same LLM with no file at all (F1 = 0.431), while structured MCP tool access achieves F1 = 0.717. This demonstrates that tool structure provides a qualitatively different mode of access that the LLM cannot replicate by reading raw syntax. The system ships as a single binary under the MIT licence.