Taxonomies

In this recording, I sit down with Clifford Lynch, Executive Director of CNI (the Coalition for Networked Information) to discuss a range of library related topics including digital preservation, cultural memory organizations, process patents, social computing, computational linguistics, and beyond.

If you haven't seen it yet, take a look at flickr. It's a photo website that is demonstrating just how wrong the nay-sayers were when they said that metadata-rich web websites would never take off because of the high cost of creating metadata. The existence of a ranking system, called "interestingness" encourages users to assign keywords to their images. Keywords are used to find the images and if images can't be found they can't be viewed and thus won't accumulate views or user comments. If accurate assigning of keywords leads to users not only finding images but finding images they like, this is encouraged by enabling users to add users whose photographs they like as contacts, at those be notified of new uploads. This again drives the accumulation views and user comments.

The blog entry below highlights one librarian's take on folksonomies ... an interesting read.  For whatever it is worth, it looks like he may work for a library system vendor (talis). 

http://wonderfulworldofmrc.blogspot.com/2005/08/folksonomies-fascinate-me.html

Personally, I don't see why multiple techniques can't compliment (maybe even propel) one another.  I've stated that before and to some degree we're beginning to test this directly with drupal.  We preloaded our database with terms from our own controlled vocabulary and the AJAX support provided by drupal provides an opportunity to focus users on those items.  As we go forward, it will be interesting to see how the use of synonyms, combined with the prospects of recommending related tags, etc. may affect classification ... it will also be interesting to see how/if we can extend that into other areas.  Hopefully some analysis of items going into the folksonomy might shed light on how they might make their way into a controlled, hierarchical taxonomy.  We might have to expand the scope of our current activities for that to become meaningful, but in the meantime, I remain fascinated by the prospects of seeing where new trends and technologies can take us. 

"Will Google’s Keyword Searching Eliminate the Need for LC Cataloging and Classification?" Well, Thomas Mann has some ideas.  Read more at http://www.guild2910.org/searching.htm

IMHO, I think it is too early to tell what the web has in store for us ...  especially while social computing is at a such an infantile stage.  As I ponder the future of interesting activities like Odeo's distributed folksonomy support, where anyone can tag an item, as well as more established, but still young players like wikipedia, metabrainz and others, I can't help but think that there is much that we can't predict about the potential for disruptive technologies ... especially if someone were to bind traditional metadata stores from sources like libraries (content that is more or less widely available),  with a more contemporary twist.  That would certainly be a fun project to work on!

A couple more podcast finds worth relaying ...

• An interview with Jimmy Wales, Founder, Wikipedia.org
• Clay Shirky, on of my favorite bloggers, on Ontologies;
  from the O'Reilly Emerging Technology Conference
  

Okay, I just ran across one of the most interesting things that I've seen in a while ... Google Suggest. Check it out. I'm thinking that it would be ever so interesting to combine a predefined taxonomy and a pre-existing list of tags as the source for something like this for tagging new blog entries and/or any other piece of content thtat should be classified. Could help reduce the use of redundant synonyms? Could it help take ethnoclassification to the next level?  It sure would be interesting to find out, and you can be sure, that if I get the time, I'll try to investigate further how we might employ this type of idea here at EDUCAUSE.