Knowledge graphs: how are facts connected?
A knowledge graph describes things and the connections between them.
What is it?
A knowledge graph records things (people, places, concepts) and the links between them, such as “Ljubljana is the capital of Slovenia”. An ontology defines what kinds of things and links exist at all. Together they let a computer reason along explicit connections.
An everyday example
When a search engine shows a box with a famous person’s date of birth and relatives, it often draws on a knowledge graph.
Why does it matter?
Knowledge graphs can be checked and corrected. Combined with language models, they can reduce made-up answers.
A common misconception
That a knowledge graph is the same as a language model. A graph stores explicit links; a model learns similarities between words.
In more detail
What is a graph?
A network of points and links, for example “person — works for — organisation”.
What is an ontology?
An agreed set of concept types and relationships, so that data can be interpreted consistently.
How does it help AI?
With linking data, searching, checking relationships and explaining facts.
Are knowledge graphs LLMs?
No. They can work on their own or as an extra source for a language model.
Remember: Connections between facts are not the same as words merely looking alike.
Reliable sources
Last reviewed 1 November 2026. An explainer, not professional, legal or medical advice. For rules and figures that change, check the official version.