Ontology Mapping Workshop¶
Facilitated workshop — instantiates Schema Conflict Resolution
A working session where domain experts lay their two formal models side by side — entities, relations, and boundary rules — and decide whether to merge, translate, or keep them federated.
Formal models rarely reveal their real structure until their authors sit across a table. Ontology Mapping Workshop is that table: a facilitated session in which the experts behind two structured models — knowledge graphs, domain ontologies, controlled vocabularies with explicit relations — elicit each model fully and then choose how the two should relate. Its defining move is that it produces two things a document alone can't: the elicited structure of each model as its authors actually understand it (including the relations and boundary rules that were never written down), and a committed integration decision — merge, translate, federate, or leave separate — made in the room by the people with authority to make it.
Example¶
Two natural-history museums merge their specimen databases, each backed by a collection ontology built over decades. In the workshop, the curators project both models. The first treats Specimen as the root entity, with Taxon, Locality, and Collector as related entities. The second makes AccessionLot the root — a batch acquired together — with individual specimens hanging beneath it. Laid side by side, the mismatch is structural, not lexical: one museum's atomic unit is the specimen, the other's is the lot. Merging into a single hierarchy would force the lot-based museum to invent per-specimen records it never kept. The experts walk their relations and boundary rules, confirm the two roots genuinely differ, and make the integration decision explicitly: federate — keep both ontologies live behind a shared query layer rather than collapse them — because unification would fabricate provenance data that doesn't exist. The workshop's outputs are the two fully-elicited models and that federate decision; the crosswalk and translation rules are built afterward from them.
How it works¶
- Elicit each model in full. Not a term list — the entities, the relations between them, the boundary and inference rules, and the purpose behind the structure, as the model's authors understand it.
- Project side by side. Align the two structures visually so mismatches in shape (different roots, different granularity, different relation semantics) become obvious.
- Surface the structural divergences. Focus on where the models organize reality differently, not merely where they name it differently.
- Decide the mode. Commit to merge, translate, federate, layer, or keep-separate — and record why — with the authority-holders present so the decision sticks.
Tuning parameters¶
- Expert coverage — whether every affected model has a knowledgeable, empowered representative present. A missing author guarantees an under-specified model and a fragile decision.
- Formalism depth — how rigorously the models are elicited (informal diagrams vs. formal axioms). Deeper formalism catches subtle relation mismatches but slows the room and narrows who can participate.
- Decision scope — one global integration mode versus per-region decisions (merge here, federate there). Per-region decisions fit reality better but multiply what must be maintained.
- Authority-in-the-room — whether attendees can actually commit their side. Without it the workshop only recommends, and the decision reopens later.
When it helps, and when it misleads¶
Its strength is catching structural incompatibility that vocabulary-level tools miss — two models can share every label and still organize the world around different root entities. It also produces a decision, not just a diagram, and having the authority-holders present is what stops that decision from unraveling afterward. The field of ontology matching exists precisely because this alignment is hard enough to warrant dedicated methods and tools.[n1]
Its failure mode is over-integration by consensus energy: a workshop that has invested a day together feels pressure to declare a clean merge, even when federation or coexistence would preserve more meaning. The classic misuse is letting the more prestigious or better-resourced model's authors drive toward absorbing the other, erasing the weaker model's distinctions in the name of a "single source of truth." The guarding discipline is to hold "keep them separate" as a legitimate, respectable outcome and to test any proposed merge against the distinctions the losing model would have to give up.
How it implements the components¶
schema_a— the first model is elicited in full: entities, relations, boundary rules, and purpose, not just its labels.schema_b— the second model is elicited to the same depth so the two can be compared as structures.integration_decision— the workshop commits to a resolution mode (merge / translate / federate / keep separate) with the authority-holders present.
It stops at the elicited models and the mode decision; it does not build the correspondence rows (category_correspondence, schema_crosswalk) — that is Schema Crosswalk Table — nor test the resolution against cases (validation_case_set) — that is Case-Based Mapping Test.
Related¶
- Instantiates: Schema Conflict Resolution — the elicitation-and-decision session that fixes which resolution mode the rest of the work will build.
- Sibling mechanisms: Boundary-Spanner Review Session · Case-Based Mapping Test · Data Mapping Specification · Glossary Alignment Table · Schema Crosswalk Table · Semantic Interoperability Review · Taxonomy Reconciliation Review · Translation Register
Editorial Notes¶
Form Classification¶
Form family: Communication, Facilitation & Learning
Rationale: The mechanism convenes domain experts to externalize two private formal models, compare them side by side, and build shared understanding before choosing a relationship.
Nearest alternative: Decision, Gate & Allocation — The group eventually chooses merge, translate, or federation, but expert-facilitated mutual interpretation is the load-bearing form.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Computer Science & Software Engineering
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Ontology Mapping Workshop is most directly rooted in computer science and software engineering's formal and practical treatment of computation, interfaces, data, and reliable systems. The lineage fits its defining practice: A working session where domain experts lay their two formal models side by side — entities, relations, and boundary rules — and decide whether to merge, translate, or keep them federated.
Related originating lineages:
- Library & Information Science — Ontology Mapping Workshop also draws materially on library and information science's traditions of classification, metadata, retrieval, preservation, and archival stewardship, which shaped this mechanism rather than merely adopting it as an application.
- Philosophy — Ontology Mapping Workshop also draws materially on philosophy's traditions of conceptual analysis, ontology, logic, epistemology, and normative reasoning, which shaped this mechanism rather than merely adopting it as an application.
Review resolution: Both independent reviews agree on primary origin computer_science; reconciliation resolves alternate_origin_disagreement, encyclopedia_synthesis_disagreement. Formative alternate lineages retained: library_information_science, philosophy. The broader reach of later applications is kept separate as domain_reach=multi_domain; origin_mode=cross_disciplinary_synthesis records how the formative lineages relate. Confidence is conservatively reconciled to medium, and encyclopedia_synthesis=true preserves the reviewers' boundary judgment.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; medium confidence.
Notes¶
Ontology Mapping Workshop and Taxonomy Reconciliation Review both end in an integration decision, but they work on different structures: this workshop compares entity-relation models with rich relations and inference, while the reconciliation review compares hierarchical or faceted category systems. When the conflict is about how entities relate, run the workshop; when it is about where categories sit in a tree, run the review.
[n1] Ontology matching (a.k.a. ontology alignment) is the research field concerned with finding correspondences between the entities of different ontologies; the annual OAEI evaluation exists because doing it well — especially for structural, not just lexical, matches — is a genuinely hard problem. ↩