Recent readings in the Open Ed course all discussed metadata and its importance to the usability of OER.
The OECD report pointed about some of the challenges: “Adding metadata to a resource is time-consuming and faces the same problem software programmers do – the person adding metadata does not know the circumstances under which people will use the resource, i.e. the search for the resource may be done from a perspective totally different from what the person adding the metadata expected, so that it will be difficult or impossible to find the resource.”
The Hewlett report talked more about metadata, including the Sematic Web, RDFs, and folksonomies.
For those looking for a more information on this in an extremely readable format, I’d suggest the book Everything in Miscellaneous by David Weinberger. He covers the whole gamut of traditional indexing strategies and their problems all the way up to the semantic web and folksonomies.
Before reading this book, my thoughts about metadata for OER mostly centered on the very large task of tagging resources by the many dimensions of learning, such as subject area, content standards (many millions, I’d guess), language, readability level, learning style, etc. I had the idea that if everything was tagged this way, a learner could use some kind of system to subscribe to RSS feeds of content that would be ideally suited to their individual needs. Differentiated learning at its finest.
After reading the book, I wonder if this would all be a waste of time. There are so many ways to tag and so many uses of materials (intended and unintended) that designers can’t begin to anticipate them all. Maybe using on-the-fly folksonomies in which everyone tags as they go for their own purposes is the simplest and most elegant solution.
The KISS principle (Keep it simple, stupid) is often a good design guideline.