IMS Vocabulary Definition EXchange (VDEX) is a public draft.
Wilbert Kraan, CETIS staff
September 09, 2003

IMS have just released a rather handy, small and general purpose specification for the exchange of vocabularies. There are a good many potential uses for the spec, but one of the main ones is as a means to facilitate interoperability between different application profiles of existing IMS and other learning technology specifications.

The VDEX spec provides a way to organise and exchange vocabularies: at its simplest, lists of terms to be used when describing something. The idea behind that is that it becomes much easier for both humans and machines to understand what something is, if it described by a known list of descriptors from an agreed source.

Even better than a flat list is a vocabulary that has a hierarchical structure- a list that not only indicates that "poodle" is a recognised term to describe an animal, for example, but also that "poodle" is an animal of the "dog" category within the "animal" vocabulary. Better still is vocabulary that expresses even more relations between descriptors; e.g. that "pooch" is occasionally used as a non-standard equivalent of the recognised term "dog" in the "animal" vocabulary.

All of these relations can be expressed directly or indirectly in IMS VDEX.

Outside of fun exercises for linguistics undergraduates, the main application of vocabularies in learning technology is to facilitate the tailoring of existing, global standards to the needs of a particular community, i.e. application profiling.

The basis for the UK LOM Core application profile, for example, is the IEEE LOM metadata standard. The LOM itself can be used to record most anything you could possibly use to describe a learning object, so that people can find it. Provided that they not only know all the possible elements of the LOM, but also the vocabulary of descriptors that go into these elements.

One of the main jobs of profiles like the UK LOM Core is the provision of a recognised UK specific vocabulary. So for a LOM element like "context" (which is in "educational") the UK LOM Core gives "sixth form college" as one of nine possible descriptors. Slap that on a learning object and every educator in the UK, and, more importantly, every UK LOM Core compliant machine, knows unambiguously that the learning object is meant to be used in a sixth form college.

Were the UK LOM Core to be defined in VDEX (it isn't yet- VDEX has only just come out as a public draft), every machine outside of the UK that can process VDEX could have a grasp of what "sixth form college" means too. Someone could even define a VDEX thesaurus that maps UK LOM Core vocabularies to the vocabularies in, say, CanCore so that UK learning objects would automatically show up in Canadian digital libraries with the appropriate descriptors.

The use of VDEX is not limited to metadata in the LOM sense, however. IMS Enterprise, for example, has a "group" element that tends to be defined in highly community specific ways. In UK Further Education (FE), for example, people are often be grouped by funding agency rather than course. Take the list of funding agencies and some human language descriptions of what they are, and how they are related to courses and qualifications, munge the lot in a VDEX file, put it online, and every human and/or machine that comes across a piece of UK FE Enterprise data would have a fighting chance of knowing what to do with it.

The public draft and associated documents can be found on the VDEX page on the IMS website.

The IMS website also has a press release on VDEX.