On Information Retrieval and Tagging – Sometimes the counter point is great. Such is the case with Gene Smith’s post on Why Tagging Is Expensive which asks the question of if the sum of the cost of tagging and wading through bad tags is greater than the benefit. Actually Gene’s piece is a follow up on another article by Ian
If you have ever used a document management system that required meta data entry on save, then you feel my pain. Being prompted for required meta data is ridiculous and I will freely admit that if I have a simple idea before lunch just knowing the stupid-human-tricks I will have to perform to create the document will prevent me from creating it. Additionally adding keywords such as tags requires something of a knowledge matter expert to properly categorize things. No, I am not siding with the librarians and their outdated hierarchies, but neither can I accurately categorize a picture of a whale (are they warm blooded? hmmmm…..) as well as a George on Seinfeld.
Ease of use changes behavior. So forcing tagging on all content authors is not the answer. And from a guy obsessed with distributed authoring, anything you do that slows down the creation of distributed content reduces the intelligence of the group. The crowd is less wise – which is bad.
As for our Tendenci software, we are trying to divide the tagging into a prompt after content add. So the user should see a keyword cloud, but then select from that on what they feel is the most relevant as the tags. And be able to add other tags free form. Having worked with SEO for years I can honestly say that people are *usually* surprised at the density of terms in their text until they are trained to think about it. This is not surprising – you must learn to be a communicator and good communication skills do not come easily. Why should you not also need tools to help you with accurate tagging? The problem might be the tags, but then again it might be, it probably IS, a problem with the content!