Cooperative categorization with folksonomies
Flickr and del.icio.us have popularized the use of folksonomies. Now Technorati has adopted the use of tags to join content from Flickr, del.icio.us and Technorati indexed blogs into one single location. Excellent job!
Folksonomies defined according to the original author Thomas Vander Wal:
Taxonomy is from "taxis" and "nomos" (from Greek). Taxis means classification. Nomos (or nomia) means management. Folk is people (from German). So folksonomy means people's classification management.
Simplest description of folksonomies are simply classification, categorization or defining keywords. The concept is not new: for a long time keywords have been assigned to web sites in meta descriptions and in controlled vocabularies.
Originally metadata has been in hands of professionals who use complicated vocabularies and tools for describing information assets. In some systems, the authors have had the opportunity to tag their content with a controlled vocabulary like IEEE LOM (learning object metadata).
The fundamental shift is the move towards an information system where the content is not tagged for personal use only but for a community to collaboratively describe information assets without any formal training in the subject. This enables wider adoption but the trade-off is that the taxonomy is not as accurate as it is in controlled vocabularies.
At first as you see Mac, Macintosh, Macs and Apple describing almost the same thing, the method seems more like chaos than control. The true benefit lies in the accidental browsing capabilities, as you find information you might never find otherwise. This enables a layer above the information to browse numerous information resources quickly and effectively, while still maintaining a level of cheer luck in finding new and interesting things.
Google is using masses to understand the content. They have abandoned the use of user provided metadata, according to their Director of Search Quality. This is because as long as there is a market to use the metadata to make money, people will abuse it (spam, fooling people to look after their commercial offerings etc). Instead, they are using the masses to find out relevant data.
When we look at folksonomies, the framework is using masses to help browsing in a similar way as Google does. Merholz notes that folksonomies could be used to create controlled vocabularies. Maybe this results in generally more useful vocabularities for personal use? This might also fix the problem with educational XML standards which have their problems according to many experts in the field.
For more information and a great analysis of folksonomies, see Grassroots Cooperative Categorization Of Digital Content Assets: Folksonomies, What They Are, Why They Work and Folksonomies - Cooperative Classification and Communication Through Shared Metadata.
It seems that in many things it's neither total control nor total chaos but the true capabilities lie somewhere in between.