Flagship: FIBO Business Entities, 2023Q3 → 2024Q3¶

owlcompare diffing two published quarterly releases of the Financial Industry Business Ontology (FIBO), a production multi-stakeholder ontology for financial-industry vocabulary, published by the Enterprise Data Management Council (EDM Council) and standardized by the Object Management Group (OMG).

Source: FIBO at EDM Council · License: MIT · Module: FIBO-BE (Business Entities) · File: OwnershipAndControl/Executives.rdf · Versions: 2023Q3 (released September 2023) → 2024Q3 (released September 2024) · Run on: 2026-06-12 · owlcompare version: 0.1.0

Why this comparison: FIBO is a real production ontology under continuous multi-stakeholder development. Academic research has documented that roughly one-third of FIBO entities change per quarter. This flagship shows what owlcompare makes of a year of evolution in one FIBO file — and the answer is a clean, comprehensible story rather than a wall of diff noise.

The headline isn't how many changes — it's what they mean

owlcompare distills 214 raw triple changes into 41 structured events, and 34 of those 41 turn out to be a single coordinated refactor: FIBO-BE adopting the OMG Commons Ontology Library in place of its own Foundations Parties vocabulary (fibo-fnd-pty-pty: → cmns-pts:). owlcompare describes a multi-stakeholder migration as a handful of human-readable changes — and EDM Council's own release notes confirm exactly that migration (see cross-validation below).

At a glance¶

Metric	Value
Total Layer 1 changes	41
Breaking changes	28
Additive changes (new classes/properties)	0
Non-breaking changes (restriction relaxations)	12
Info changes (labels, comments, metadata)	1
Renames detected	0
Anonymous structure changes	0
Datatype facet changes	0
Unexplained Layer 0	10

The Unexplained Layer 0 line is owlcompare's headline correctness claim. Here all 10 are meaningful ontology-level metadata — owl:imports reorganization, owl:versionIRI bump, and cmns-av:copyright year update — not decoder noise. There is nothing in this diff owlcompare failed to account for except changes that have no higher-level structure to fold into. (More in complete coverage of the diff.)

This window contains zero renames and zero anonymous-structure changes — no unionOf shifts, datatype facets, or dcterms:isReplacedBy assertions. That is characteristic of this editorial period (a vocabulary migration, not a restructuring), not a gap in owlcompare; the rename detector and Component 12.5's anonymous-structure decoder simply had nothing to fire on here.

Source files¶

The exact FIBO files used in this diff are committed to the owlcompare repo, ready for anyone to fork or rerun:

2023Q3 baseline: examples/fibo_demo/v1/ (FIBO-BE module, git tag master_2023Q3)
2024Q3 target: examples/fibo_demo/v2/ (FIBO-BE module, git tag master_2024Q3)
License: MIT, preserved at LICENSE-FIBO

The specific file diffed in this showcase is OwnershipAndControl/Executives.rdf. The Reproduce this diff section below has the one command that turns these files back into every output format on this page.

Cross-validating with EDM Council's release notes¶

This is the showcase's unique value-add: owlcompare's findings cross-referenced against EDM Council's own published release notes for the intervening quarters. Every quote below is verbatim and maps to a specific change owlcompare detected.

From the EDM Council 2023Q4 release notes — the migration begins¶

"Changes made to Business Entities in Q4 included minor spelling corrections and elimination of duplicate content with respect to the new Commons ontologies as well as elimination of deprecated elements from early 2023."

— EDM Council, FIBO 2023Q4 Release Notes

The same notes forecast the specific work owlcompare later detects: "We will continue this work in Q1 2024 for roles and parties." That "roles and parties" alignment is precisely the migration that lands inside this diff.

From the EDM Council 2024Q1 release notes — the migration lands in BE¶

"Revisions to Business Entities in Q1 were primarily with respect to support for migration of certain FIBO FND ontologies to the OMG's Commons Ontology Library."

— EDM Council, FIBO 2024Q1 Release Notes

owlcompare reports the corresponding changes — 34 of the 41 total: 10 entities re-parented from fibo-fnd-pty-pty: to cmns-pts:, and 24 restrictions migrated from fibo-fnd-pty-rl:isPlayedBy to cmns-rlcmp:isPlayedBy:

🔴 Class reparented: fibo-be-oac-exec:AuthorizingParty: fibo-fnd-pty-pty:Actor → cmns-pts:Actor (lateral) 🔴 Restriction added on fibo-be-oac-exec:Executive: exactly 1 fibo-be-le-lp:LegallyCompetentNaturalPerson cmns-rlcmp:isPlayedBy

From the EDM Council 2024Q1 release notes — which vocabularies moved¶

"The primary revisions to Foundations in Q1 reflect migration of several ontologies to the OMG's Commons Ontology Library. These include the entire Roles ontology, much of the Parties ontology…"

— EDM Council, FIBO 2024Q1 Release Notes

This explains why the namespaces changed: the Parties ontology moving to Commons is why fibo-fnd-pty-pty:Actor/PartyInRole became cmns-pts:Actor/ PartyRole (the reparents and the elects/nominates domain/range shifts), and the Roles ontology moving is why fibo-fnd-pty-rl:isPlayedBy became cmns-rlcmp:isPlayedBy (the 24 restriction changes).

FIBO validates itself, inside the diff

owlcompare also surfaces a skos:changeNote that FIBO's curators embedded directly in Executives.rdf for this very release: "…modified to replace content that is now available in the OMG Commons Ontology Library (Commons) v1.1 (FND-380)." The diff is self-validating: owlcompare detected the migration mechanically, and FIBO's own change note — ticket FND-380 — confirms it.

The full interactive report¶

Open the interactive HTML report ↗

Screenshot of the FIBO diff HTML report

Self-contained HTML (~370 KB); opens offline; no dependencies. Also available as Markdown (rendered on GitHub), JSON, or JUnit XML.

How owlcompare describes the changes¶

Four patterns from this diff, each shown as owlcompare's literal output followed by what it means.

1. Lateral reparenting — the Commons migration, consolidated¶

owlcompare detects FIBO's Commons-adoption migration as 10 lateral reparents, consolidated from 20 raw triple changes (each reparent is an rdfs:subClassOf/ rdfs:subPropertyOf removed and added at the Layer 0 level):

🔴 Class reparented: fibo-be-oac-exec:AuthorizingParty: fibo-fnd-pty-pty:Actor → cmns-pts:Actor (lateral) 🔴 Property fibo-be-oac-exec:authorizes reparented: fibo-fnd-pty-pty:actsOn → cmns-pts:actsOn (lateral)

Rather than show you 20 disconnected add/remove triples, owlcompare folds each into one statement and tags the direction (lateral) — it recognizes the new parent is a same-named concept in a different namespace, neither a generalization nor a specialization. Seven properties (authorizes, hasAuthorizingParty, isAuthorizedBy, …) and three classes (Authorization, AuthorizingParty, ResponsibleParty) all moved this way.

2. Severity classification at scale¶

The 24 restriction changes split cleanly into 12 breaking additions and 12 non-breaking removals. Here is one entity, fibo-be-oac-exec:Executive, on both sides of the migration:

🔴 Restriction added on fibo-be-oac-exec:Executive: exactly 1 fibo-be-le-lp:LegallyCompetentNaturalPerson cmns-rlcmp:isPlayedBy 🟡 Restriction removed from fibo-be-oac-exec:Executive: exactly 1 fibo-be-le-lp:LegallyCompetentNaturalPerson fibo-fnd-rel-rel:hasIdentity

owlcompare's severity rule is asymmetric and conservative: adding a constraint (red) can invalidate data that was previously valid → breaking; removing one (yellow) only relaxes the model → non-breaking. That asymmetry is what lets a reviewer trust the breaking count as a worst-case signal.

A v2 opportunity, stated honestly

These 24 changes are really one migration of the isPlayedBy predicate. A future owlcompare version could collapse such pairs into a single restriction_predicate_migrated kind. For v1 they are correctly reported as independent breaking/non-breaking changes — the axioms genuinely differ at owl:onProperty, so owlcompare does not pretend to a semantic equivalence it hasn't proven.

3. Signature evolution¶

The migration also shifts property signatures — domains and ranges — and owlcompare reports each as its own structured change:

🔴 Domain changed on fibo-be-oac-exec:elects: fibo-fnd-pty-pty:PartyInRole → cmns-pts:PartyRole 🔴 Range changed on fibo-be-oac-exec:elects: fibo-fnd-pty-pty:PartyInRole → cmns-pts:PartyRole

elects had both its domain and range retargeted to the Commons PartyRole concept; nominates likewise, and hasResponsibility's domain moved to cmns-pts:Party. These are exactly the kind of signature shifts that quietly break downstream SPARQL and SHACL if a reviewer misses them.

4. Complete coverage of the diff¶

owlcompare accounts for every change in this diff. Of the 214 raw Layer 0 triple changes, all but 10 fold into one of the 41 structured Layer 1 events shown above. The remaining 10 are ontology-level metadata that have no higher-level structure to fold into:

🟡 Removed: …Executives/ owl:imports <https://spec.edmcouncil.org/fibo/ontology/FND/Par… 🟡 Added: …Executives/ owl:imports <https://www.omg.org/spec/Commons/PartiesAndSituat…

The import reorganization toward Commons, the owl:versionIRI bump from the 2023 to the 2024 IRI, and the cmns-av:copyright year update — meaningful Layer 0 metadata that doesn't have a Layer 1 equivalent.

One change in the diff uses owlcompare's complex_class_expression_changed kind:

🔴 Complex class expression on fibo-be-oac-exec:BoardCapacity changed (deep)

This appears once across the entire 41-change diff and signals that a deeply-nested anonymous class expression on BoardCapacity changed in a way that the v1 structured decoder reports with its outer structure rather than a fully-unfolded breakdown.

Reproduce this diff¶

The exact FIBO source files and reproduction command are committed to the owlcompare repo:

# Clone owlcompare
git clone https://github.com/Ajala111/owlcompare.git
cd owlcompare

# Install owlcompare
uv sync                     # if you have uv
# OR: pip install -e .       # if you prefer pip

# Run the flagship diff
uv run python -m owlcompare diff \
  examples/fibo_demo/v1/OwnershipAndControl/Executives.rdf \
  examples/fibo_demo/v2/OwnershipAndControl/Executives.rdf \
  --format html --out flagship-report.html

# Open flagship-report.html in your browser

Or regenerate every output format used on this page:

python scripts/generate_flagship.py

The FIBO source files in this repo are unmodified copies of EDM Council's publicly-released versions (git tags master_2023Q3 and master_2024Q3; MIT license — see examples/fibo_demo/LICENSE-FIBO). The reproduction output may differ slightly from this page if you run a newer owlcompare version with updated classifiers.

Attribution¶

The Financial Industry Business Ontology is published by the Enterprise Data Management Council (EDM Council) and standardized by the Object Management Group (OMG). FIBO is licensed under the MIT License (Copyright © 2020 Enterprise Data Management Council); a copy is preserved in this repo at examples/fibo_demo/LICENSE-FIBO.

This showcase demonstrates owlcompare's diff capabilities. owlcompare is not affiliated with EDM Council, OMG, or any FIBO contributor. We use the publicly-available FIBO files as a representative test case for the kinds of multi-stakeholder ontology evolution owlcompare is designed for.

owlcompare is authored by Olatunji Felix Ajala (phelz), who chose FIBO as the flagship subject for its permissive license, its rapid quarterly evolution, and its broad recognizability to the ontology-engineering community.

← Back to owlcompare docs