I2IDL Releases Open Semantic Knowledge Infrastructure for Digital Learning
I2IDL has released version 0.0.89 of the I2IDL Glossary of Digital Learning, extending the glossary's role as an open reference resource and a machine-readable knowledge graph for digital learning.
The glossary currently includes 397 concepts connected through curated semantic relationships, 31 collections, 541 evidence records, and links to external vocabularies. Built using JSON-LD, SKOS, Schema.org, Dublin Core Terms, and PROV-O, the underlying model provides persistent identifiers, source attribution, provenance, and structured relationships among concepts.
With the addition of a public REST API and Model Context Protocol (MCP) server, developers and compatible AI systems can search, retrieve, and explore the glossary's curated knowledge directly. A read-only SPARQL endpoint supports semantic queries, while the Knowledge Graph Explorer provides an interactive environment for investigating connections, comparing concepts, and discovering relevant scholarly literature.
These services create opportunities for integrating common terminology into AI-assisted research, educational software, standards development, higher education courses, and other modalities and systems that depend on consistent definitions and traceable sources. And provenance has been our target all along. To that extent, all machine-access interfaces are read-only, preserving the distinction between curated knowledge and externally generated interpretations. The terms and concepts themselves are immutable and expressed as RDF with full versioning embedded.
I2IDL continues to develop this infrastructure as an openly accessible resource for the digital learning community, with an emphasis on interoperability, transparent provenance, and long-term stewardship.
We invite you to explore the resources:
We still consider ourselves in the prototype phase, and therefore invite your comments. Note that new capabilities and features are planned and we have a number of sources that we intend to include in future iterations of the glossary. If you have open CC BY material you would like us to consider, please get in touch.
One of the more interesting things on the roadmap has to do with getting more people involved in consensus (and debate) around definitions. We have purposely built the underlying data model to support this, and we think our approach may offer a new way to think about how to engage with reference materials. Details forthcoming.
