Usually people who are fans of taxonomies do not get along with folksonomy fans. Our speaker, Tom Reamy, made sure that we knew going in that he was on the taxonomist’s side of things before he started talking. The funny thing was the next thing he said was that Library Scientists and Folksonomists (or believers in folksonomies) get along even worse that taxonomists and folksonomists – coming straight from Internet Librarian where everything was about social software and social tagging – I found this funny (I do know that the people at Internet Librarian were from one subset of the library science group – and that they’re more likely to like new tagging options than those who weren’t there).
So are you wondering what a folksonomy is? A folksonomy is a classification done by the user or some amateur – not a professional cataloger. Tom mentions that in addition to being data added by the user it’s also a social experience – you can see what’s popular by visiting del.icio.us and viewing the tag cloud.
Tom moved on to give us some advantages & disadvantages (all of which I think were from Wikipedia – but I’m not 100% sure).
- Simple – no learning of formal classifications or heirarchy needed
- Lower cost of categorization (using the users)
- Open-ended – can respond quickly to changes
- Relevance – users are closer to the data and so they can tag things better than an expert
- Multiple Dimensions – put together by “communities” of like-minded taggers
- Easy to tag any object
- They don’t work very well. They’re great for personal indexing, but they aren’t great when someone else wants to find something.
- Don’t compare favorably with controlled vocabulary
- No structure – “onomy” should not be part of the name – “onomy” = structure
- Get caught up in jargon
- Subject matter experts are not professionals (I ask you to read what Dave Snowden said in his keynote – and then David Weinberger [not written by me yet])
- No quality control – all based on popularity
- and many more…
Tom’s argument is that the 2 don’t compare – they are radically different – I can probably think of a few people who would disagree – but they’d probably be folksonomy fans.
So what does Complexity Theory have to do with this all? While complexity theory does relate to maths & sciences, the key concept is that it is self-organizing.
An example of a complex system is ant tunnels. The ant doesn’t know what a tunnel is, it doesn’t know why it needs to make tunnels, it just knows that there is something in the way and it needs to be moved. In this case evolution has lead to organization. Other features of complexity theory are feedback, local interactions and large number elements.
Tom goes on to bring this all back to our intranets by making the point that our intranets will not evolve into a structure – we need to put that structure there (something I did by writing my own wiki application instead of using one that depends on links for structure). Intranets are also a great place to use tagging because it’s a controlled environment and everyone is tagging within the same context (this is especially true for a library intranet where all of the librarians are also familiar with taxonomies).
So, in order to make tagging more effective Tom feels we need some changes. We need (most importantly) feedback – a way to rank tags and some rules that say if this tag is ranked poorly it’s removed from the list of tags – so this is feedback with consequences. We need to allow only certain people to tag – once again get feedback – all for ranking of taggers to see who’s using helpful terms in their tagging. Lastly, we need to use a combo of explicit (people) and implicit (software) ranking methods to get better tags and more useful search results.
Overall, a very interesting talk – and it’s even more interesting after you hear about some of the speakers who followed (which I better get to writing about).