I’ve done a number of workshops demo-ing how to search various learning object “repositories” and invariably deal with the question, “Why don’t we just do a Google search?”.
Strangely, having built one sort of similar system myself, I am asking the same question.
Stephen Downes today shared the announcement of the Commonwealth of Learning’s Learning Object Repository being released, “An online database of learning content that provides software to Commonweath countries free of charge” and the software was being made available as well.. sounds interesting enough to click around.
Hitting the technical documentation, you get alphabet soup explanations:
The COL Learning Object Repository (or in short COL LOR) integrates eRIB and pakXchange such that the local repository of eRIB is disabled and replaced with pakXchange, and pakXchange is modified to act as an EduSource node for the purpose of searching.
Easy for you to say… what the heck is all that?
So then I thought, give the thing a whirl and see what I can figure this does. Now I am looking at the eRIB “eduSource Repository-In-A-Box” (Yikes, hope that fares better then the long gone “Web Course In a Box!”).
But apparently it lets you do “federated” searches, or searches for learning objects across multiple sites. I tried some real simple queries to generates lots of results. Unfortunately, not taking the cue from Google in that search results produce retrievable URLs, I cannot easily link you to what I saw. But I ran three searches on:
* volcano
* economics
* heart
Hoping to see some learning objects, what I found is that 95% of the results are simply links to web pages, many of them course syllabi [1] [2], in some cases images [1], pages not found [1] [2] [3], this page has moved [1] [2], a thesaurus [1]
Yes, an unscientific sampling, and perhaps I misunderstand the purpose of this site is more to demonstrate the search technology than the content.
But if these are the learning objects that are meant be “reusable” chunks of content, I am totally mystified as to what this giant piece of technology has created- a keyword search engine that finds web sites, and pretty meta data for web sites (many of which do not exist any more).
If that is the case, we ought to just use Google, eh?
I posted the following in the Teaching and Developing Online blog on October 7 2004.
Nice to see we are all thinking along the same lines.
Learning Object Repositories are a thing of the past as seen through the cyber glasses.
In my humble opinion, the number one issue of using learning objects to create course material is one of saving time (searching vs. starting from scratch). The challenge to find just the right learning object to meet instructional requirements is almost as great as developing a standardized globally acceptable meta-system. These issues alone are why the learning object repositories concept has not taken off.
Most experts when trying to explain why the repository system has become bogged down will try to blame it on a lack of meta-tagging or standards. The development of a universally accepted meta-tag system would only result in a lot of work for undergraduate students who would be needed to apply these to the existing learning objects. The expertˆ‚s solution to this costly method of meta-tagging the learning objects is to get the users of the repositories to tag the objects as they search and sort through them. This user meta-tagging system is flawed because even the easiest user applicable tagging system will require time and expertise that most searchers who are pressed for time even to find an object that comes close to matching the style and needs of the course do not have. The other choice is to get the developer at the time of creation to tag the object, again flawed because the developer has barely the time needed to adequately create the learning object never mind meta-tagging so someone else out there on the web would be able to find it and use it. A robot type tagging system has the best chance of success. But once tagged the fact still remains that the majority of learning objects are created by course designers to fit into a specific location and style. This limits the usefulness of much of the material which would be found within a repository and greatly reducing the chances of finding a sufficient number of objects that look like they belong together and fit the needs. After taking all this into consideration perhaps it is better just to create the material from scratch. Then once the course has been created, enhance and support it with robot tagged learning objects that you might (I stress the word might) have time to search for once you have taught the course for a few years. Which comes to another question, has anyone every ever really completed an online course? Every online course that I have written never seems to be done; I am always finding new exciting learning objects to support my created content. Maybe this is the answer to how and when learning objects should be used in creating learning material.
Learning object repositories are a thing of the past. The material should reside in the location it was originally created for. Once tagged in that location it can be found as easily there as it could be found in a repository. The money being spent collecting all the objects into a repository should be spent on developing an automated tagging system then the complete web becomes the repository, removing the need for duplication of objects.
Possibly a narrow tinted view through the cyber glasses.
Yan Simard and I had a discussion on this back in January through my posts in ICTlogy.net and his comments to them. Not heavy literature but some musings around the subject. You can find them in the following links:
http://www.ictlogy.net/index.php?p=208
http://www.ictlogy.net/index.php?p=210
http://www.ictlogy.net/index.php?p=219
The repository is of course an artifact of the LCMS model where you did not simply link to resources, you ‘acquired’ them, loaded them to your own system, and then embedded them into content packages.
No browser can read a content package, of course, but content packages are intended for shipping and delivery – they are intended to be run on stand-alone systems, such as CD-ROM or internet-disabled intranets.
The nominal reason for this is that a course designer cannot be sure that a resource will remain on a third party site, and must therefore procure an actual copy of the resource. The larger reason is that content packages are viewed in the manner of, say, books or even software programs, and the entire content must be ‘republished’ as a whole. Following this publishing model, of course, it also ensures that any content used in a course is cleared and paid for.
Challenge content packaging and repositories – as I have – and you will hear no end of stories about unreliable content, limited connectivity, and lack of control over design and format.
In my opinion, online learning and CBT should never have ‘merged’ to become e-learning — what works on CD-ROMs doesn’t work on the internet, and vice versa. And we are now saddled with a cumbersome system of federated search, content packages and learning design, none of which came into being with any of the affordances of the internet in mind.
— Stephen
I agree with your thoughts here Alan. Most of our LORs, including the CAREO instance I helped build and populate, are long on metadata and short on objects. They are more like a social bookmarking site without many of the cool features of Furl or del.icio.us
What’s most outrageous about these parasites are the millions in tax dollars that this unwieldy system has cost, compounded by the news that its manifest pointlessness has claimed another victim.
The learning object mafia must be stopped, and Leon Lighips is just the man to do it. I’ve got one of these blogs now, and it’s already the only weblog worth reading.
You smug and self-satisfied ed tech bloggers had better not rest too easy. You are for me, or against me. You can be my partner, you can be my employee, or you can be my victim. I can be your friend and protector. I can be your doom. The choice is yours. Only you can make it.
Choose wisely my friends.
To continue this thinking along the same lines, it sure seems that we were creating a big white elephant. I never could get my brain around this stuff, as it didn’t make business sense. The closest that I can find to true reusible objects are the permalinks that blogs produce. Produce once, find often. Glad to see many sane opinions on this site.