elearning2.0

Graham Attwell has just asked for some help sorting out 'repositories' in his head, from the point of view of project-funded researchers who want to publicly share the work which they post on project sites.

I have sort of tracked the repositories debate over the last couple of years but have lost sense of where it is going. Now i need to sort out a couple of questions in my head. It seems to me one of the problems is that repositories have always been envisaged as institution wide or cross institution things.

I'm not completely sure where Graham is going with this, since I know him to be aware of lightweight social software tools as well as heavier-weight educational materials repositories, digital libraries and content management systems. I am going to interpret that Graham's line of pondering went something like :

1. Is there an easy way for ordinary researchers to share their documents with - and make them discoverable by - the wider world?
2. Is there a 'best way' to do that, different from the 'easy way'?
3. Where is the development of educational or research 'repositories' going, and if it is not dead, is this the way to do [1]?
4. What is the relationship between the lightweight 'social bookmarking' services and the more controlled and featureful repositories and digital libraries?
5. Is there a middle way between ad-hoc tagging and controlled metadata vocabularies, so that us mortals can tag our materials to make them discoverable?
6. What about content management systems and tools such as Plone and KNotes?

I'll try to say a bit about each of those questions below, but my main reasons for replying are to point out just two things: (a) a way to approximate [1] now using very simple and widely available tools, and (b) work that I've started planning which should make more sense of [4, 5 and 6] above.

Extended text for this entry:

Answers in brief

1 - the simple way to make my resources discoverable

Every researcher should make themselves familiar with the great tools and services for 'social bookmarking' which have become available in the past few years. For examples, see my own del.icio.us service bookmarks tagged as follows: social-bookmarking, social-citation and social-content (perhaps popping up into the larger space of my social bookmark tags, where there are many related materials noted). The social-bookmarking tag at connotea.org is also a useful place to start, and it is worth studying Social bookmarking - Wikipedia, the free encyclopedia. For a sense of what the whole wide world is tagging on that topic, see the social-bookmarking tag from all del.icio.us users. Three key things to be aware of - all of the social bookmarking and social citation services have open APIs so that they cam be 'mashed up' with other sites, tools and services; they all have extremely rich RSS news-feeds available for all the content and tagging; because of their open-ness, popularity, ubiquity and simplicity, a large and growing number of tools is emerging to leverage and enrich their user experience (By the way, these are usually taken as the defining qualities of a web2.0 service - but there's no space here to go into that).

It is possible to make use of services like del.icio.us very easily to tag your own materials in a big public service and thus make them as discoverable as your tags are predictable by the people who look for resources. I heartily recommend doing so - the effort spent now will certainly be proved well worth it as time goes by ( and all of the services make it simple to extract, export, import, and migrate :o)

To start using social bookmarking to 'make' a repository, just register at one or more of the services, use the cool tools they provide to bookmark your web-posted materials, and add the tags you would expect your 'audience' to look for.

As Graham has clearly guessed, herein lies the problem with the simplicity of 'folksonomies' - how does your audience know what tags to look for? In the simple-but-popular folksonomies, tags are just strings of text which you make up as you go along (see also categorising/folksonomy). Unlike controlled-vocabulary repositories or metadata schemas, you can use any tags you like, and your 'audience' does not in general have any way of finding out what tags have been used for materials like those they want to find. remarkably, this has not been a roadblock to massive uptake, since in many cases people are more than capable of mutual mindreading - of using terms which simply make sense to other people. There are certainly problems with this tag-soup chaos (problems which I and others are scheming to overcome with new services in the future), but it is still well worth posting and tagging your own materials and resources.

It is worth noting that the sociability of these services can help you to find out what tags to use in order to 'meet up' with other people's tagging practices. One of the great merits of posting bookmarks for your own materials in the big public services is that other people might have bookmarked the same links. if they have, the goodness of co-tagging kicks in: you can view listings of all the other users who have bookmarked that item, and see (a) what other items they have bookmarked and (b) what tags they use.

I have found that the social bookmarking and social citation services are particularly useful for sharing materials with an audience that is bigger than my immediate co-workers but smaller than the whole wide web of users. Once you have tagged a pile of stuff, you find that you have a lot of useful links to share with people. I've been trying to demonstrate that above, by using links from my own bookmarking :o)

Of course, all of the above is useless if you do not already have your materials on the web somewhere publicly accessible where the hosts do not mind the bandwidth implications of telling the world about them. If you do not have your materials on the web, or the materials are inaccessible to the public, or they are heavy-bandwidth consumers, you'll need to make use of one of the content management systems, or better yet one of the emerging crop of Web2.0 file hosting and social content services (eg flickr for photos, see also social-music, social-video, services/bit-torrents etc.

Is there a 'best way' to do that, different from the 'easy way'?

You might encounter heated debate about the merits of ad-hoc tagging on open but commercial systems, as opposed to using controlled classification systems in public digital libraries or content repositories. The advantage of using controlled 'vocabularies' is that both tagger and seeker can see what tags there are. The disadvantage - and this is a big disadvantage in my experience is that the taggers cannot express themselves freely as they come upon interesting new resources and uses for them. Generally, controlled vocabulary systems will also provide tools to speed the tagging process - this is good for taggers who need to make use of gigantic taxonomies but bad for entry-level users who only need a hundred or so categories. There are also arguments on either side of the validation issue - "who says?" - ie can a seeker rely on the items tagged such-and-such being proper examples of such-and-such materials (if this sounds ridiculously paranoid, think of all the spam in your email inbox). The social-bookmarking approach is silent on this question, but because of the open APIs you can count on 3rd party tools emerging to deal with tag-spamming if it becomes a big problem. If you require more stringent validation of metadata quality, there are schemes which might be applied in the future - these are after all just a refinement of 'social' (see webtech/metadata).

There is no room here for me to go into the issue of what is the best approach for various purposes, but I would like to take the opportunity to point out that they are not mutually exclusive, and to stringly encourage researchers to at least use the open, simple social-bookmarking approach for all their materials. If other approaches are known to you, then by all means use those as well - think of it as being like advertising in print and on the radio.

Where is the development of educational or research 'repositories' going?

I cannot answer that question offhand. I 'track' several ongoing projects in that area via their RSS feeds but have not recently distilled an overall picture of the field. Graham's question has provoked me into a bit of thought on the issue. I've struck a couple new delicious tags for beginning to assemble links to relevant materials: resources/open-libraries and webtech/repositories. There are a lot of projects that seem to cleave into Digital Library projects and Learning-Object Repository projects (for digital libraries see List of digital library projects - Wikipedia, the free encyclopedia, also projects such as Connexions - Sharing Knowledge and Building Communities).

What is the relationship between the lightweight 'social bookmarking' services and the more controlled and featureful repositories and digital libraries?

Again - a good question, and I wish I knew the answer. I would like to think that any sensible architect of learning materials repositories is currently rethinking everything in the light of the remarkable success of what I call the "Small Loose Standards" Way ( see my tags standards/small-loose and microformats and my draft report on Standards and Architectures for the SIGOSSEE project for more on the small-loose way :O) My feeling is that the Capital R "Object" Repositories beloved of systems designers of the old fashioned IMS school are rapidly losing currency in higher education, but - bizarrely - gaining credibility among decision-makers in the schools sector. Service oriented architectures and large-scale open-access digital libraries will almost certainly change the way we all think about repositories. Because so little usable functionality has made its way from the 'R' folks into the hands of real-world teachers and researchers, I would guess that real-world users will vote with their mouse-buttons and start creating their own repositories ad-hoc in order to explore great new learning methods and experiments (but that some sysadmin 'standards' fascists will try to close them down for using unsanctioned tagging :O)

In the medium term, I am guessing that there will be a lot of development at what has been the boundaries between 'CMS' and 'Repository' and 'bookmarking service', and that this development will be given momentum from Open Source and Open Standards advocates but also officially supported by enlightened authorities like JISC and the NZ government.

Is there a middle way between ad-hoc tagging and controlled metadata vocabularies?

Not yet, but there should be. I am about to start writing up our own proposed work to support such middleweight/middle-way tagging. I was deeply involved for many years in the metadata for learning materials game; in the course of those projects we learned a lot of valuable lessons about what users really need in order to share resources. In the past few years, working in open-source e-community portals development, we've learned a lot more. And in the past year or so we have intensively explored the lightweight 'small loose' systems.

I wish there was time or space here to go into details about our plans, but there isn't, so the following hints will have to suffice for now :O)

We propose to do work on three fronts to support 'repositories' that are usable by real working communities:

Tagging tools.

We have learned that there is a real need for high-quality tagging assistant client software; once you have 400 or more del.icio.us tags for instance, all the existing tools start to swoon. I have a lot of experience writing tagging tools. The services all have public APIs. Roll on taggingtools.com (watch that domain in 2006)

Services to grow and share your own loosely-controlled vocabularies.

This could be a very important part of creating repositories for the masses: A way to make up our own tags in realtime without having to put up with total chaos. A way to negotiate shareable, controlled vocabularies and taxonomies without having to wait for standards-organisation timeframes. The simple idea is to create tools to make it as easy as possible for groups of users to propose, discuss and agree vocabulary and taxonomy items, while also letting their vocabularies grow as they tag. We have considerable experience in trying to write such systems, so we know they are tricky - but we also know they are do-able. Thanks to the advent of the popular social bookmarking services, and of open-source community portal platforms, it will be a lot easier to do this now than it was in 1999 when I first tried it! So.... roll on two more domains to watch in 2006: sharemytags.com and grouptags.com. (I'll try to post an elaboration of how this will work as soon as possible - there is more to it than I've had space for here.

Integration-with and proxying-by community portals.

As popular as the new social software services are, most people still do not know about them, nor do they want to try tracking this exploding area of services and companies. And as 'social' as these services are, their scope is either personal or global - which makes it tricky to use in a group or community-wise way (an example: the almost all the resources tagged 'counselling' in del.icio.us are about psychological counselling, whereas to our NGRF members, 'counselling' means careers counselling. We have begun exploring the communal use of the big public social software services on behalf of community portals, and are beginning to scope out an interesting space of functionalities and requirements. Watch the NGRF and SIGOSSEE sites as flagships for our efforts in that direction in 2006, for instance shared citable-objects bases backending the 2000 and growing annotated references in the NGRF.

Technorati tags: blogoposium1, web2.0, repositories