IRI scheme for all the Artifacts

Here’s the first part of the IRI scheme for all artifacts:

Base IRI Scheme

The Base IRI is the same for all FIBO products that are published on :


Ontology IRI Scheme

Here’s the continued IRI scheme for the ontology artifacts:


Vocabulary IRI Scheme

Here’s the continued IRI scheme for the vocabulary artifacts:

/<artifact>[.<format>[.<compression>]] IRI Scheme

Here’s the continued IRI scheme for the artifacts:


UML IRI Scheme

Here’s the continued IRI scheme for the uml artifacts:


Documentation IRI Scheme

Here’s the continued IRI scheme for the doc artifacts:


Proposed layout for the ontology documentation

The documentation “site” could actually be a shadow of the other product IRI schemes, for instance one OWL class IRI could be backed by a documentation page as follows:

The above OWL class IRI could be represented in the documentation as follows:

Static files IRI scheme

Here’s the continued IRI scheme for the static artifacts:



Element Description
protocol Is always https, we do NOT support http, please be aware of the fact that the difference between using http and https, even though its only one letter, is in many cases significant. Two OWL axioms that only differ with this single letter are actually not the same thing.
host Is always although the OMG ratified version of FIBO will be (also) published on the OMG site but they have as slightly different IRI scheme
family Is always fibo (lowercase), it could be that the EDM Council will publish other product lines like this as well such as DCAM
branch or tag In case of a git branch name like pink or BE-123 just use the exact same name in the exact same “case”. In case of a tag, when the tag starts with v then remove the v, so v1.2.3 becomes 1.2.3
domain A top level domain, same as the directory names in the root of the fibo repository, such as be, fnd and so forth.
sub-domain A sub-domain as it is used (exactly) as a directory name at the second level in the repo’s directory hierarchy. So this must be UpperCamelCase.
sub-sub-domain Same as sub-domain but then at the 3rd level. In the ontology product this field is optional.
artifact The meaning of the field artifact changes per product. In the ontology product this is the actual file name, without extension, of the OWL ontology (assuming that we only have 1 ontology per file). See the various artifact tables below for all possible values.
artifact-type For static files, this is image for all images, javascript for all Javascript and css for all CSS stylesheets.
format For RDF files: the file extension representing the serialization format, see the table of file extensions below. For any other files this is just the standard file extension like .jpg and .gif
compression The compression encoding that should be used by the server, currently only gz is supported, see also the paragraph Accept-encoding below

HTTP Request headers

Accept header

When <format> (and therefore also the <compression> part) are not part of the URL then the HTTP headers can be used to specify which format needs to be returned by the server. The so-called “Accept Request-header” can specify a list of MIME types, where each MIME type corresponds with one of the supported formats, for instance if you want the RDF/XML version of an ontology:

Then specify the following HTTP header:

Accept: application/rdf+xml

If you want the same file as Turtle, specify this:

Accept: text/turtle

See the paragraph “File extensions & MIME Types” below for a listing of the supported values of the Accept header (see column MIME Type)


If you want a compressed version of the artifact then you can do that either by specifying the full file name and adding the .gz extension, or by specifying the Accept-encoding request-header as follows:

Accept-encoding: gzip

File extensions & MIME Types

Extension MIME Type Description
rdf application/rdf+xml or application/xml RDF/XML format
owl Not supported via accept header OWL/XML format
ttl text/turtle or application/x-turtle Turtle format
nt application/n-triples N-Triples format
n3 text/n3 or text/rdf+n3 N3/Notation 3 format
jsonld application/ld+json JSON-LD format
nq application/n-quads N-Quads format