New projects often introduce new terms which can sometimes be used in different ways between team members. Especial when there are multiple teams. This article will clarify the meaning for the following terms and help explain how they are related to one another.
Most importantly I hope this will help you use the correct term at the right time and help correct someone who may be abusing a term. We need to start off with a fundamental concept.
At the most basic level, a graph is a connection of two things. This graph has something connected to something else. RDF is a way we can create these graphs. I won’t go too much into how RDF describes data except it can store the graph as a triple in the form
(subject, predict, object).
At this point we don’t know what each node is or what the meaning of the connection between the nodes are. The graph could be:
(person, married to, person) (concept, related to, concept) (course, teaches about, concept)
Notice how I use the triple notation to explain the graph. RDF gives us a general way to define the parts of a graph, but is there a way we could describe what the nodes and the connection mean in a graph. In other words, is there a way we could define the semantics of the graph.
What is an Ontology
An ontology defines what and how things are connected together.
(classes, are connected by properties to other, classes)
Returning to our earlier list of things, we can see that the subject and objects are classes and the predicates are the properties.
(person, married, person) (concept, related, concept) (course, teaches, concept)
The best way to see how ontologies work is with a few examples.
A Social Ontology could define the following classes and properties. A graph of this ontology would only have those elements.
An education institution could define different meanings to the nodes and connections in the graph.
This graph could look like
Skos is a popular ontology. This ontology has concepts that can be related to each other in broader and narrower terms. There a more classes and properties than I show, the important thing is this one ontology can be used to create hierarchies and a lot of other things.
The project I’m on has a lot of new terms. Hopefully now we know the difference between an
ontology and a
SKOS:Concept. There is a lot more to talk about and this should lay the foundation of a common vocabular of terms.