Graph fundamentals
Five guides that assume you have never built a graph
What a graph model is, what an ontology adds to one, how property graphs and semantic graphs differ, and which query language belongs to which. No product in any of them. They are here because the decisions further on are hard to make without this vocabulary.
The set
Read in any order, though the first one carries the vocabulary
Between fifteen minutes and half an hour each. They are introductions rather than references, so they stop before the edge cases.
Graph models
Graph structures, graph models, and how both differ from the relational model you already have. Covers the model types, whether you can mix them, and how to decide what data belongs in a graph at all.
read graph models
Ontologies explained
What an ontology is, how it differs from a schema, which standard ones are worth reusing before you write your own, and what changes about a graph once it has one.
read ontologies
Property graphs
Nodes, edges and properties. Why labelled property graphs became the popular default, and what they give up in exchange for that flexibility.
read property graphs
Semantic graphs and RDF
Triples, RDF and the semantic web stack. Best read straight after the property graph guide, because the difference between the two is the decision most projects reach first.
read semantic graphs
Graph query languages
What a graph query language is, how Cypher, SPARQL and Gremlin differ in shape rather than in syntax, and where GQL standardisation leaves all three.
read query languages
Choosing a graph database
Not part of this set, but the page most people want next. Fifteen databases, which query languages each speaks, and what that means for a model you have already built.
go to graph databases
Also worth your time
Three things that are not guides
The old hub mixed these in with the guides above. They are a different kind of reading and they are easier to find when they are not.
The blog
Release notes, modelling arguments and walk-throughs, including how to build a knowledge graph without coding and how to convert several joined tables into RDF.
go to the blog
The KGC23 talk
Twenty minutes from the Knowledge Graph Conference in New York. Building a graph end to end on stage, from source data to a loaded database.
watch graph.build.done
The lifecycle
Once the vocabulary is in place, this is the shape of the work: the eight steps of building a knowledge graph and which of them can be automated.
go to how to build a knowledge graph
Next
When the vocabulary stops being the hard part
The guides are deliberately product-free. If you have read enough of them to know what you want to build, the fastest next step is seeing it built on your own domain.
- 45 minutes, your data, no installation
- You keep whatever we model
- No obligation to buy anything
