Notes on " When tags work and when they don't: Amazon and LibraryThing"

There is an excellent posting at the LibraryThing's blog on a study on tagging, it compares tags in LibraryThing and in Amazon.com. What is also a great reading are the comments, there are many of them which are very deep and reflective, too.

Lots of issues in this posting touch upon what we are currently thinking in MELT, where we are implementing a social tagging tool for teachers to tag digital learning material. A few points that I will elaborate upon later:

• Tag categories (factual, subjective, personal); how there are so many opinion tags in Amazon compared with LT.
  
• The idea of separate "buckets" vs. "works" (LT)
  
• How the tagging interface works and supports the workflow (to show tags from others or not? how does this effect on convergence of tagging?)
• Tags are not reviews or ratings! they serve different purpose.
  
• Reasons and drivers for tagging
• Number of tags to become important
• Why implement tags in the first place, what is there for Amazon? they already have all sorts of ways to recommend content, do they still need tags? it's just another feature on top of all other whistles, it's not part of the design for the site to work.