What is a knowledge graph?
A knowledge graph stores knowledge as entities and the named relationships between them, so software can follow those relationships to answer questions.
How it works
Each fact is a node linked to another node by a relation with a name: a city is the capital of a country, a person works for a company. Standards such as W3C’s RDF write these facts as subject, predicate and object.
Google popularised the term in 2012 for the graph behind its search panels. Teams also use the phrase loosely for any map of how their documents and ideas connect, which is a graph of documents rather than of facts.
An example
A team’s documents form a graph when the spec links to the decision that shaped it, the decision links to the research it relied on, and each link says why: depends on, inspired by, blocks.
Common mistakes
- Calling any tag cloud a knowledge graph: without named relations, it is a list.
- Drawing a graph nobody maintains, so it shows how things were connected a year ago.
In Nodary
Nodary draws a project’s documents as a map: each document is a star, folders gather them, and links carry a reason (related, inspired by, blocks, depends on) that reads the right way from either end. People and agents follow the same links.
What Nodary does not do: it is not an RDF store and has no query language over facts: its graph links documents, not the statements inside them.
Try it on your own documents. The Free plan needs no card.
Create a free workspace