Skip to content

Knowledge graphs

Knowledge graphs are data structures that represent knowledge as named entities (nodes) linked together via named relationships (edges). They are graphs combined with semantics.

Knowledge graphs generally adhere to a data model, or ontology, which specifies the different types of nodes and edges, their properties and possible relations. In that sense, there is an infinity of different knowledge graphs for one given ontology.

Knowledge graphs can be built automatically using existing data, and large companies use them to improve their search engines and recommendations.

However, we are interested to take another route, and ask: what if we defined knowledge graphs ourselves? Crowdsourcing the creation and curation not only of nodes but also of their relations and properties?

In broad terms, we could call those collaborative, co-produced, or co-modelled knowledge graphs. If democratic mechanisms were to be rooted in the edit process, why not calling them democratic knowledge graphs1?

While knowledge graphs generally imply a fixed view of a topic, the process of building a knowledge graph collaboratively or democratically is inherently dynamic.

  1. We've found only this PhD thesis which used this term, referring to the Wikidata platform. ↩