October 30, 2007
Comments on a Typology of LORs
I couldn't find a blog for Rory McGreal, author of a draft article called A Typology of Learning Object Repositories, so I guess I'll just post my comments here and hope he looks for backlinks. Except the article is identified with a handle and not a URL, so I don't know if he'll find anyway. Oh well. Maybe if someone who reads this blog and knows Rory and email him the link or something.
Anyway, this is an article that analyses some learning object repositories and creates a typology, pretty much as it says in the title. It covers a range of fairly well known repositories such as CAREO and ARIADNE etc., and provides comparisons of capabilities as well as how many resources each repository manages.
In this respect I think the article is a useful survey, and the tables are quite handy to see what different systems are doing. I would have liked some other categories, however: for example, which LORs support RSS feeds for new items (called "Alert" in the IMS Digital Repositories Interoperability reference model, which though not referenced actually describes the other functions used in the paper), and which ones offer a search API (SRU for example) and not just a search form. It also misses out JORUM, which is the JISC-supported repository for "Learning Objects" in Higher Education in the UK.
Another aspect, raised by D'Arcy is that some of these repositories are actually zombies, with no further funding and no-one actively maintaining the servers.
However, I don't think Rory's analysis of Learning Objects holds up too well. Partly this is a problem of categories. If "Learning Object" is really taken to mean any digital resource used for learning, then clearly they are wonderfully successful as the whole Internet is stuffed with them. On the other hand, if we take "Learning Object" to mean an IMS-packaged, modular, context-free, LOM-described chunk of learning lego, then the picture is very different. This is what I'm referring to in the presentation of mine that Rory cites - not OpenCourseWare PDFs, not Wikipedia entries, and not web pages or blog posts used for learning. However these are some of the things counted in the works cited in the analysis as "Learning Objects".
Perhaps just as we had to distinguish between "Web Services" (the service model that uses WSDL and SOAP, plus any number of WS-* extensions) and "web services" (any machine-usable service based on the web, including REST, web 2.0 APIs, AND Web Services) it would be useful in any of these discussions to make a distinction between (uppercase) "Learning Objects" and (lowercase) "learning objects", or quite frankly any assertion pro or con can be supported.