domains / data / ontology-mapping / ontology-mapper

ontology-mapper

Skill card

combined

The two skills share the name 'ontology-mapper' but target completely different domains and procedures. The first maps construction data to building ontologies (IFC, COBie, Uniclass) using a Python class-based approach. The second maps materials science data (crystal structures, space groups) to scientific ontologies (CMSO, ASMO) using a script-based CLI approach. They do not overlap in content, inputs, or outputs, so crowning one would hide a distinct, valid alternative for a different domain.

Repositories

Other copies

Combined skill

# Ontology Mapper

## Overview

This skill maps domain-specific data to standard ontologies. The specific domain and target ontologies depend on the context:

- **Construction Domain**: Maps construction data to IFC, COBie, Uniclass, OmniClass, MasterFormat, and UniFormat.
- **Materials Science Domain**: Maps materials science terms, crystal structures, and sample descriptions to ontologies like CMSO and ASMO.

## Workflow

1. **Identify Domain**: Determine whether the input data is construction-related or materials-science-related.
2. **Select Target Ontology**: Choose the appropriate standard ontology for the domain.
3. **Map Concepts**: 
   - For construction: Use the `OntologyMapper` class to map fields to ontology concepts.
   - For materials science: Use the appropriate script (`concept_mapper.py`, `crystal_mapper.py`, or `sample_annotator.py`) to map terms or parameters.
4. **Validate**: Check for validation warnings and unmapped fields.
5. **Report**: Generate a mapping report with coverage and confidence levels.

## Conflicts

- **LeoYeAI/openclaw-master-skills**: Focuses on construction ontologies (IFC, COBie, etc.) and uses a Python class-based approach.
- **FreedomIntelligence/OpenClaw-Medical-Skills**: Focuses on materials science ontologies (CMSO, ASMO) and uses a script-based CLI approach.

## Notes

- The construction skill requires `python3` and uses a dataclass-based model.
- The materials science skill requires `python3.8+` and uses standard library only, with external JSON config files for ontology mappings.