view printer-friendly version (opens in new window)
ADL to make a 'repository SCORM'
Having safely delivered the SCORM 2004 (aka SCORM 1.3) e-learning content format, ADL outlined its plans for the Content Object Repository Discovery and Resolution Architecture (CORDRA) at the first ADL international plugfest in Zurich, Switserland. Like SCORM, the lovely named CORDRA will not be a new specification, but a reference model that combines a number of existing standards and specifications.
In ADL technical director Phil Dodds' words, the forthcoming reference model is motivated by an identified need for contextualised learning object discovery. The ideal is that learners would be able to discover and identify relevant material from within the context of a particular learning activity. There's to be no need to leave a familiar environment, and the results should be "more precise for our environment" than what can be done using web search engines or other discovery technologies.
The 'more precise' here is meant as discovery results that give a specific solution to a specific skills gap; searches should not just give results that give a wider overview of, or introduction to a particular field. Instead results are to be targeted to a particular need in a given context.
The example given by Mr. Dodds was of an engineer aboard an aircraft carrier that is about to receive a plane with a piece of malfunctioning avionics. The engineer is familiar with the system that needs repairing, but is unfamiliar with the particular version. A search from within the ship's VLE should give him the exact learning object that she needs to bridge the knowledge she has of the general system with the knowledge she needs to repair the specific instance in the plane.
The motivation for ADL to step into the field of repositories now is, according to Phil Dodds, similar to the situation in the e-learning content field when ADL started work on the Sharable Content Object Reference Model (SCORM): "we're seeing lots of interesting developments, and now it's a matter of tying them all together". The one difference with the SCORM development is that there will be no need to develop parts of the reference model by ADL itself this time around. There are enough specifications and standards in the digital repository field to combine and adapt to ADL's needs.
Which specifications and standards will be used is not yet fully determined, but there is a prime candidate for solving the identifier question.
This issue revolves around the problem of being able to uniquely identify a known learning object. Once systems can be certain that an identifier refers to exactly the intended object, a lot other of other discovery problems (such as determining location, the right metadata record and so on) becomes tractable. Bearing in mind a number of learning object peculiarities such as the fact they can be combined and taken apart and re-used, the problem is much like the one addressed by ISBN numbers in the book world.
There are a number of schemes around today that address the problem of uniquely identifying digital objects, but the CORDRA is most likely to use the Handle system. This is a system that ultimately resolves to a registry held by the Corporation for National Research Initiatives (CNRI) in the US. They appoint registrars in particular communities. The idea, then, is that a learning object's existence is announced within a community's registry, and that the registry itself is known to the root registry at CNRI.
One particular use case that ADL would like Handle to address are the auxiliary resources that are part of SCORM 2004. Auxiliary resources are a generic term for any material that can be used as reference or 'scaffolding' material alongside the SCO (SCORM learning object) based instruction. Such resources are often maintained independently of the learning material, and it is a real hassle to synchronise what's in the learning object with what's the latest version, of, say, a manual. To get around the issue, SCORM 2004 objects could contain Handles that resolve to resources, rather than the resources themselves. Discovering the best version of a resource -depending on the context and learner- would be dealt with when the learning object is run, rather than when it was designed.
Some of the input that has already gone into CORDRA came from the ADL academic co-lab in Madison, Wisconsin. They held a major, Hewlett funded conference about repositories last year, and have published a number of follow up resources since. Most of the work so far on the CORDRA itself has been done by the ADL Joint Forces co-lab in Orlando, Florida, however. They conducted a large scale requirements survey among the US forces and have been building test implementations. They are also collecting use cases, including those from outside the military.
As the work still has some way to go, it is not known when any normative documents will come out.
The ADL Academic co-lab has a page with repository resources, including a database of projects, a wiki and a report of the summit.
The ADL Joint co-lab has released version 1.0 of the DoD Repositories Use Cases
More about the Handle system is available on the Handle webpages
A video stream of Phil Dodds presentation will be available on the ADL site shortly, as will slides of a presentation on the subject by Steve Slosser of the ADL Joint co-lab. Dan Rehak's presentation on research for the US Army into discovery and resolution of auxiliary resources in SCORM 2004 is available from the Carnegie Mellon University Learning Systems Architecture Lab webpages.