Walker, E. (2003). Commentary on Oleg Liber and Bill Olivier, Learning Technology Interoperability Standards, Chapter 12 of: Reusing Online Resources: A Sustainable Approach to eLearning , (Ed.) Allison Littlejohn. Kogan Page, London. ISBN 0749439491. [ www.reusing.info ]. Journal of Interactive Media in Education , 2003 (1) Special Issue on Reusing Online Resources. ISSN:1365-893X [ www-jime.open.ac.uk/2003/1/ ].

Chapter 12: Learning Technology Interoperability Standards

Oleg Liber and Bill Olivier

Commentary by Ed Walker

IMS Global Learning Consortium, Inc.

Summary Comments

Bill Olivier and Oleg Liber have done a service to the field of on-line learning by beginning this chapter with an succinct summary of the benefits of interoperable media and software for education and training, as well as a concise differentiation of specifications, reference models, and standards. At the time this chapter was written, most work on specifications and standards concentrated on data exchange. The data consisted of items of content, learner characteristics, or administrative information. To be sure, such data is exchanged in learning interactions. But while they are necessary, data specifications by themselves are by no means sufficient to characterize the technical requirements of the interactions.

The distinction the authors draw between de jure and de facto standards foreshadows their central critical observation; which in simple terms, is to speculate whether the notes and scales of standards are powerful enough to represent the music of learning. As they put it,

"In our opinion, the key question now facing eLearning standards development is how they can ensure support for the broadest range of pedagogic approaches."

Indeed, work on a second generation of specifications that deals with the behavior, as well as the structure, of learning resources, had begun at the time the chapter was written. Olivier and Liber quote Professor Rob Koper's rationale for the now completed work on specifying learning activities in the IMS Learning Design specification.

"So far IMS has specifications to describe learning (Meta-data), package learning (Content Packaging), to support enrolments on courses and get results back (Enterprise), to test learning (Question and Test) and to support the exchange of information about learners (Learner Information), but there is nothing to describe and define the learning process itself. That is what EML brings to IMS Learning Design."

Certainly much work remains to be done to refine and extend specifications for the means which underlie effective learning interactions. The notes and scales we have do not yet constitute a fully developed and enduring system for notating learning. However, primitive as they are, using them have begun to reveal that use cases (examples) and profiles (applied standards) which differentiate pedagogies on the basis of their technical requirements -- as opposed to merely on the features of one pedagogical technique or another -- are difficult to find and understand.

Are the specifications inadequate because they instantiate out-moded or inadequate pedagogy or because more modern or more adequate pedagogical approaches have yet to be expressed in terms from which specifications can be developed? In fact, both the means for delivering learning interactions and the goals of those interactions are immature. The answer to the either-or question is, not surprisingly, "yes".

Maturing Means

The status of the specifications reviewed has changed since Olivier and Liber's summary of the state of the art was made. New specifications that were yet to be started at the time of their review have been added. Existing specifications have been revised in response to user feedback and technical innovation. Many of the specifications which Olivier and Liber mention have experienced widespread adoption and use.

One duly accredited international standard has been approved, the Learning Objects Metadata standard of the IEEE/LTSC. A content management standard based on the AICC CMI is about to emerge from the IEEE balloting process. Similar progress is being made in the ISO SC36 Working Groups.

Work also continues in IMS, CEN/ISSS, SIF, and other specification development organizations. In IMS, new work on the selection and ordering of instructional content has grown from the Learning Design and Simple Sequencing specifications and from the assessment oriented Question and Test specification. A specification defining Competencies, or educational objectives, has been released, and accessibility extensions to the learner information package and meta-data are under development. The Enterprise specification is being extended and is now complemented by a Digital Repositories Interoperability specification. The DRI specification is serving as the basis for a growing collaborative effort between the learning technologists and the digital library community to address the interaction between learning environments and distributed repositories.

The MIT Open Knowledge Initiative project has just released specifications for Application Programming Interfaces (APIs) that are aimed at providing interoperability between learning environments and implementations of system services. This work is the result of a multi-institutional project funded by the Mellon Foundation, and its results already are being applied in developments at several universities worldwide.

CANCORE and SINGCORE specify meta-data for the education and training communities in Canada and Singapore, respectively. The Schools Interoperability Framework Consortium is in the process of profiling the IMS Question and Test specification for assessment applications in the K-12 or schools community, and the SIF Framework and Zone Integration Server are serving as the basis for data aggregation and exchange in several communities worldwide. In the UK, the Office of the e-Envoy is developing an e-Government Interoperability Framework which includes a suite of specifications for education and training.

The surest sign that adoption is progressing is rising demand for tests of compliance and for maintenance and evolution of the specifications and standards. Both kinds of demand have appeared and are being met. The ADL Co-Laboratory has developed and deployed test suites for SCORM. SIF has developed tests for the SIF Zone Integration Server. IMS has initiated an International Conformance Program in order to facilitate the development of application profiles, test suites, and a program of certification which can be adapted for use world-wide

It seems clear that the specifications which have been developed so far are achieving the scale of adoption and momentum from use that is required to carry them beyond their origins in computer based training and their initial focus on data alone. Specification development shows signs of becoming a sustained and continuous process.

Maturing Goals?

Work to define the behaviours which constitute learning activity -- both in the microcosm and in the aggregate -- also has continued. The Learning Design specification stimulated by Koper's EML has been released, as has a Simple Sequencing specification arising from the instructional design requirements of the SCORM community. On-going work on the selection and ordering of assessment related content has progressed in the community using the IMS Question and Test Specification.

Is this progress good news or bad? Or does it merely heighten the concern that Olivier and Liber express that

"... eLearning standards will constrain Internet supported learning by freezing a sub-set of existing practices, or whether specifications can be provided that can support the development of new, enhanced, but yet to be developed approaches to learning which the Internet makes possible"?

Whether they are enabled by the internet or not, supporting different pedagogical approaches will depend in the first place on understanding their technical requirements. Could it be that the key challenge or risk for the future of on-line learning is not the provenance of the specifications or their representational power, but the development of computationally relevant descriptions of pedagogical distinctions?


The Simple Sequencing , Question and Test, and Learning Design specifications have established a basic capability for characterizing learning interactions. Is that capability sufficient to support the theoretical constructs of scholars of learning? That question can now be examined empirically.

So the field has arrived at a crux. Both the means and the ends of learning technology are immature, but they do exist. Almost certainly the specifications are inadequate. But are they sufficiently mature to create a virtuous cycle of improvement between those with a practical orientation and those with a theoretical point of view?

Previous Section Top Next Section