“janmartinkeil” — Fediverse search results on home.social

Dylan Van Assche @dylanvanassche · 2022-11-30 · 11:21 UTC

@janmartinkeil It was more a high level overview of this field, this event is to network with other fields to exchange knowledge and ideas.

Regarding your points (my view):

Ontologies are more like the description of how the data is modeled. For example, the ontology #FOAF says `foaf:Person` is a Person. While the #KnowledgeGraph (KG) will have `<Dylan> rdf:type foaf:Person` to say that I'm a person modeled according to the FOAF ontology. The data in the KG is modeled after the ontology.

#foaf #knowledgegraph

Jan Martin Keil @janmartinkeil · 2026-05-14 · 09:57 UTC

Atanas Kiryakov, President of Graphwise, in his keynote at ESWC points out a merging advantage of (re-)using public ontologies: LLMs already know these ontologies, which increases the interoperability of the data by making use of knowledge graphs as context easier and cheaper.

#ESWC2026 @eswc_conf #KnowledgeGraphs #LLM

#eswc2026 #knowledgegraphs #llm

Jan Martin Keil @[email protected] · 2026-05-14 · 09:57 UTC

Atanas Kiryakov, President of Graphwise, in his keynote at ESWC points out a merging advantage of (re-)using public ontologies: LLMs already know these ontologies, which increases the interoperability of the data by making use of knowledge graphs as context easier and cheaper.

#ESWC2026 @eswc_conf #KnowledgeGraphs #LLM

#eswc2026 #knowledgegraphs #llm

Jan Martin Keil @[email protected] · 2026-05-14 · 09:57 UTC

Atanas Kiryakov, President of Graphwise, in his keynote at ESWC points out a merging advantage of (re-)using public ontologies: LLMs already know these ontologies, which increases the interoperability of the data by making use of knowledge graphs as context easier and cheaper.

#ESWC2026 @eswc_conf #KnowledgeGraphs #LLM

#eswc2026 #knowledgegraphs #llm

Jan Martin Keil @[email protected] · 2026-05-14 · 09:57 UTC

Atanas Kiryakov, President of Graphwise, in his keynote at ESWC points out a merging advantage of (re-)using public ontologies: LLMs already know these ontologies, which increases the interoperability of the data by making use of knowledge graphs as context easier and cheaper.

#ESWC2026 @eswc_conf #KnowledgeGraphs #LLM

#llm #knowledgegraphs #eswc2026

Jan Martin Keil @[email protected] · 2026-05-14 · 09:57 UTC

Atanas Kiryakov, President of Graphwise, in his keynote at ESWC points out a merging advantage of (re-)using public ontologies: LLMs already know these ontologies, which increases the interoperability of the data by making use of knowledge graphs as context easier and cheaper.

#ESWC2026 @eswc_conf #KnowledgeGraphs #LLM

#eswc2026 #knowledgegraphs #llm

Jan Martin Keil @[email protected] · 2026-05-12 · 09:15 UTC

Already yesterday, I presented at the ESWC 2026 Pre-Conference in the Knowledge Graphs for Sustainability Workshop (KG4S) the Design-for-Circularity Ontology (D4CO). Related artifacts are available online.

🔗 Ontology: https://w3id.org/d4co
🧑‍💻 Repository: https://gitlab.com/dlr-dw/d4co
📄 Paper Pre-Print: https://elib.dlr.de/224130/

#eswc2026 @eswc_conf #KG4S @kg4s #ontologies

#eswc2026 #kg4s #ontologies

Jan Martin Keil @[email protected] · 2026-05-12 · 08:09 UTC

Recently, I was discussing about the pros and cons of different unit ontologies. However, I was not able to really express, why I am in favor of an ontology, which not only provides the direct conversion factors between units for direct application, but also provides the explanation for these values. In her keynote, Maria-Esther Vidal provided an explanation for my intuition (free after Aristotle): "We think we know a thing only when we know its cause." #eswc2026 #UnitsOfMeasurement #Ontologies

#eswc2026 #unitsofmeasurement #ontologies

Jan Martin Keil @janmartinkeil · 2026-04-28 · 12:07 UTC

LintedData is a linter for RDF and Ontologies for easy use in CI pipelines, we recently released. It checks for common violations of best practices in ontology engineering.
GitLab: https://gitlab.com/dlr-dw/linteddata/
Docker: https://hub.docker.com/r/dlrdw/linteddata/