Learning Metadata - Are We There Yet?

For nearly 10 years there has been active debate regarding the nature and application of metadata frameworks in support of Learning solutions.

At first we only had the Dublin Core, the original set of elements contained in the specification were:

Title
Creator
Subject
Description
Publisher
Contributor
Date
Type
Format
Identifier
Source
Language
Relation
Coverage
Rights

Not long after, the IMS Global Learning Consortium began building more detailed standards, specifically tailored to the Learning industry. Those specifications have proliferated and now include:

• IMS Metadata 1.3 (core specification)
• IMS Question & Test Interoperability Specification (2.1)
• IMS Tools Interoperability Guidelines 1.0
• IMS General Web Services 1.0
• Content Packaging Specification 2.0
• IMS Learner Information Package Specification 1.0.1
• IMS ePortfolio specification 1.0

There are in fact more IMS Standards, but these are the recent and major ones. There are also other standards floating around - at one point the IEEE was working towards a Learning Object Model metadata standard but that effort has now been more or less merged into the SCORM standards.

In a future post we are going to contrast these with the evolution of metadata in other industries with particular focus on those industries which already are or might converge with Learning. The main question being asked in this post however covers the larger issue - how specialized should learning standards be ? Are they too specialized already and if so why is that important?

Standards are a bit of a double-edged sword at times, while they make many thing possible they also come with a conformance price-tag. The really hard question is also determining whether or not some of the same functionality is being recreated across different standards in different industries - this comes into play for Learning in regards to content management, knowledge management and as you might have noticed from earlier in this post, SOA as well.

It is my contention that Learning integration is not limited to Learning systems and that many of these standards can be stripped away and replaced be either simpler models or by industry standards emerging in other fields. This if followed would provide our industry the mechanism to embed learning capability and culture across many enterprise domains where it is currently not considered either important or viable.

Copyright 2008, Semantech Inc.