Results Chain Ontology

The Vision

Conservation projects rely heavily on “results chains”—diagrams that map strategies to environmental impacts. While effective for individual projects, these frameworks are almost universally trapped in static PowerPoint slides, making portfolio-wide analysis impossible. You cannot query a picture.

The Solution

I built the Results Chain Ontology, an open-source OWL knowledge graph schema that translates thirty years of flat diagrams into queryable, machine-readable data.

• Core Schema: Models the fundamental flow of conservation logic (Strategies → Threat Reductions → Impacts) while enforcing logical integrity with SHACL shapes.
• Global Alignment: Automatically maps project nodes to authoritative external sources like the CMP Taxonomies, Conservation Evidence database, ENVO, and UN Sustainable Development Goals using SKOS.
• Reproducible Pipelines: Features a robust Python and Makefile ingestion pipeline that scrapes evidence databases, runs crosswalks, merges graphs, and generates human-readable HTML documentation automatically via Widoco.

The Outcome

By treating conservation frameworks as queryable data rather than static diagrams, the ontology unlocks:

• Portfolio-Level Reasoning: Funders can trace the exact environmental impact of specific strategies across thousands of grants using a single SPARQL query.
• Agent-Assisted Design: LLM agents grounded in the ontology can instantly surface proven interventions for new projects based on historical evidence.
• Interoperability: Projects instantly inherit crosswalks to every major conservation knowledge source for free.

The takeaway: A well-aligned semantic layer is the cheapest, highest-leverage intervention available to the conservation sector. The same pattern driving enterprise AI—Data → Semantics → Ontology → Agents—works beautifully across entire scientific disciplines.