experiment: ROBOT BOT-extracted slim modules for RO + IAO + CCO#26
Merged
experiment: ROBOT BOT-extracted slim modules for RO + IAO + CCO#26
Conversation
Replaces the full upstream releases of RO and IAO (and adds CCO instead
of stubs) with ROBOT-extracted slim modules using --method BOT (a syntactic
locality module variant). BFO 2020 stays as a full local import (matches
IAO's own pattern; BFO is small and ISO-standardized).
Why BOT and not full imports:
- BOT is the OBO Foundry / ODK / ROBOT-recommended default ("when in doubt,
use this one"). GO, RO's own Makefile, and ~hundreds of ODK-managed
projects ship BOT-extracted slims.
- SLME-BOT carries a formal deductive-conservativity guarantee: the
extracted module entails exactly the same things as the full upstream
for terms in the seed signature (Cuenca Grau / Horrocks / Kazakov /
Sattler 2007/2008).
- MIREOT is legacy and does NOT preserve property-typing or property-chain
axioms reliably; ARCO's gate axioms use anonymous inverse property
expressions on iao:0000136 that would silently drop from the DL fragment
if iao:0000136 ended up declared only as AnnotationProperty in a slice.
BOT preserves property typing and characteristic axioms reliably.
- Full imports of RO + IAO worked but pushed HermiT CI to 35-40 minutes
and projected to 1-3 hours when CCO was added. BOT slim modules drop
HermiT CI to ~6.6 minutes locally on the merged ontology.
What's new:
- 03_TECHNICAL_CORE/ontology/imports/iao_bot.owl (115 KB, was 610 KB)
- 03_TECHNICAL_CORE/ontology/imports/ro_bot.owl (291 KB, was 1.2 MB)
- 03_TECHNICAL_CORE/ontology/imports/cco_bot.owl (421 KB, was stubs only)
- 03_TECHNICAL_CORE/ontology/imports/seeds/{ro,iao,cco}_seed.txt
(the explicit seed signature ARCO depends on, version-controlled)
What's removed:
- 03_TECHNICAL_CORE/ontology/imports/iao.owl (full release no longer needed)
- 03_TECHNICAL_CORE/ontology/imports/ro.owl (full release no longer needed)
Pinned upstream releases (provenance):
- BFO 2020 (ISO/IEC 21838-2:2021): full local file
- RO release 2025-12-17: BOT-extracted from ro.owl
- IAO release 2026-03-30: BOT-extracted from iao.owl
- CCO release v1.7-2024-11-03: BOT-extracted from merged 11 modules
(last release before v2.0 namespace migration; preserves ARCO's existing
cco: IRI namespace http://www.ontologyrepository.com/CommonCoreOntologies/)
Verification (this branch, local):
- Pipeline: PASS, Sentinel-ID classified AnnexIII1aApplicableSystem (entailed).
7744 asserted -> 27454 post-reasoning, +19710 entailed (was +47888).
- Five-scenario regression: 5/5 PASS. Same classifications as full-import
state. Both adversarial scenarios (equivalency decoy, blank-node ghost)
still entail correctly under BOT-preserved upstream axioms.
- ROBOT validate-profile DL: PASS (881 KB merged ontology in profile).
- HermiT reasoning: ~6.6 minutes locally (was ~35-40 min on full RO+IAO+ARCO).
CI workflow:
- robot-validate.yml merge step now points at the BOT modules.
- Job timeout reduced 90 -> 30 minutes (HermiT now ~6 min, 5x headroom).
ARCO local CCO stubs in ARCO_governance_extension.ttl are intentionally
preserved. Two of the six (cco:DirectiveICE / cco:DescriptiveICE rdfs:subClassOf
iao:0000030) bridge CCO's information-content-entity hierarchy to IAO's;
upstream CCO does not link these. The other four are now redundant with
the BOT module's upstream axiomatization but cause no conflict; left in
place for now.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Replaces the "Why full upstream imports" subsection with "Why ROBOT BOT slim modules" framing that matches actual OBO Foundry / ODK convention. Updates the Foundational ontology versions table to show "ROBOT BOT-extracted slim module" rather than "Full ontology" for RO / IAO / CCO. Updates the headline claims for the new state: - Hero TL;DR bullet 1: "with ROBOT-BOT-extracted slim modules of OBO Relations Ontology, Information Artifact Ontology, and Common Core Ontologies — the OBO Foundry / ODK standard pattern, version-pinned and reproducible from seed files in the repo." - Certificate sample: ENTAILED TRIPLES ADDED 47888 -> 19710. - Getting started: 17266 -> 65154 -> 7744 -> 27454. The new "Why ROBOT BOT slim modules" subsection lists five practical reasons: formal SLME entailment-preservation guarantee, OBO Foundry / ODK convention, MIREOT being legacy and unsafe for property-typing-sensitive reasoning, seed-file reproducibility, and operational scaling. The previous "five reasons full imports are right for ARCO" framing is gone — that argument turned out not to match what mature BFO-based projects actually do (RO ships a curated BFO subset, IAO MIREOTs everything except BFO, CCO consumers typically use the modular structure rather than the full pre-integrated file). The full-import experiments are preserved in git history at PRs #24 and #25 as the empirical evidence base for the BOT decision. Documentation only; no code or ontology change. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Replaces the full upstream imports of RO and IAO (and the CCO local stubs)
with ROBOT-extracted slim modules using
--method BOT, a syntactic localitymodule variant. BFO 2020 stays full-imported (matches IAO's own pattern).
Why BOT and not full imports
BOT is the OBO Foundry / ODK / ROBOT-recommended default. Gene Ontology, the
OBO Relations Ontology itself, and the hundreds of ODK-managed projects all
ship BOT-extracted slim modules for their dependencies. SLME-BOT carries a
formal deductive-conservativity guarantee (Cuenca Grau / Horrocks / Kazakov /
Sattler 2007/2008): the extracted module entails exactly the same things as
the full upstream for terms in the seed signature.
What changes
03_TECHNICAL_CORE/ontology/imports/iao_bot.owl(115 KB, was 610 KB full)03_TECHNICAL_CORE/ontology/imports/ro_bot.owl(291 KB, was 1.2 MB full)03_TECHNICAL_CORE/ontology/imports/cco_bot.owl(421 KB, was stubs only)03_TECHNICAL_CORE/ontology/imports/seeds/{ro,iao,cco}_seed.txt(theexplicit seed signature ARCO depends on, version-controlled)
03_TECHNICAL_CORE/ontology/imports/iao.owl(full release no longer needed)03_TECHNICAL_CORE/ontology/imports/ro.owl(full release no longer needed)run_pipeline.py,test_scenarios.py, androbot-validate.ymlnow load the BOT modules; CI job timeout reduced90 -> 30 min (HermiT now ~6.6 min locally vs 35-40 min on full imports).
ROBOT BOT slim modules" — five practical reasons grounded in the formal
SLME guarantee, OBO Foundry / ODK convention, MIREOT being legacy, seed-
file reproducibility, and operational scaling.
Pinned upstream releases
2025-12-17: BOT-extracted fromro.owl2026-03-30: BOT-extracted fromiao.owlv1.7-2024-11-03: BOT-extracted from merged 11 modules(last release before v2.0 IRI-namespace migration; preserves ARCO's
existing
cco:IRI namespace)Local CCO bridging stubs (kept)
Two local assertions in
ARCO_governance_extension.ttlremain:cco:DirectiveICE rdfs:subClassOf iao:0000030cco:DescriptiveICE rdfs:subClassOf iao:0000030Upstream CCO does not link
cco:InformationContentEntityto IAO'siao:0000030. ARCO's bridges connect the two ICE hierarchies. The otherfour CCO declarations (
cco:Person rdfs:subClassOf bfo:0000040,cco:Organization rdfs:subClassOf bfo:0000027, the property declarations)are now redundant with the BOT module's upstream axioms but cause no
conflict; left in place for in-file readability.
Verification
Local:
python 03_TECHNICAL_CORE/scripts/run_pipeline.py: PASS, Sentinel-IDclassified
AnnexIII1aApplicableSystem(entailed). 7744 asserted ->27454 post-reasoning, +19710 entailed (was +47888).
python 03_TECHNICAL_CORE/scripts/test_scenarios.py: 5/5 PASS. Sameclassifications as full-import state. Both adversarial scenarios
(equivalency decoy, blank-node ghost) still entail correctly.
robot validate-profile --profile DL: PASS — merged ontology is OWL 2DL conformant.
CI (this PR): see ROBOT Validation run.
Why this is safe
The BOT extraction has a formal entailment-preservation guarantee for terms
in the seed signature. The seed signature was generated from grep of every
(bfo|ro|iao|cco):IRI in the ARCO TTL files plus defensive pulls ofro:0000056andro:0000092(which BOT would pull anyway via inverse-of /sub-property axioms). The pipeline + scenario regression confirms identical
classifications. The DL profile and HermiT consistency confirm the merged
ontology is logically well-formed.
The full-import experiments leading up to this branch are preserved in git
history at PRs #24 (combined IAO + RO promotion) and #25 (README post-merge
cleanup). They are the empirical evidence base for the BOT decision: full
imports work but cost ~30 minutes of HermiT CI time per run, and the cost
projects badly when CCO is added. BOT modules deliver identical
classifications with the formal guarantee plus convention-matching audit
story.
Test plan
run_pipeline.pyPASStest_scenarios.py5/5 PASSrobot validate-profile --profile DLPASS🤖 Generated with Claude Code