On Tags

There is much talk about tags and ESP game and much excitement about it in many blog circles.
It appears now a days that people believe that all the difficult problems that we can NOT solve today using computers because they are too difficult, can be somehow solved using tags. This is particularly true about problems related to Artificial Intelligence. Ahn’s work in ESP game is attractive to people because it gives hope that tagging problem is normally tedious and will not be solved, but using game approach it could be. Time will tell whther this results in assigning tags to all images on Google and elsewhere or not.

The issue, however, is lot deeper than that. There is a strong impedance mismatch resulting in serious difficulties in use of current information systems. As I stated in a Communications of ACM article ( http://www.ics.uci.edu/~jain/papers/CACM_Experiential_Cmputing.pdf ):

“Most conventional information environments actually work against human-machine synergy.
The human mind is very efficient at conceptual and perceptual
analysis and relatively weak at mathematical
and logical analysis; computers are the opposite. The
process of designing information environments
involves logical and mathematical principles;
humans eventually interact with these systems using
a logical approach. Even the interfaces are designed
to expect users to formulate anything but the most
obvious searches using logical combinations, thus
discouraging many users.”

Clearly today’s computer vision systems are at the level below a toddler and hence use of human help in solving image search problem, sounds a good one. I am not sure that this is such a good approach, however. Tags are very subjective and ESP game does not remove that. Remember that keywords (or tags) assigned to articles — even by authors and editors — were relatively ineffective in text search for articles in comarison to the exhaustive text processing approaches that made current search engines what they are.

I am not saying that tags as proposed in Flickr or now in Google are not useful – they clearly are. But don’t expect that a ladder — even a real tall one — will take us to moon. Tags will help in solving some simple problems, but may turn out to be distractions in the long term. We need lot more than tags to solve image search problem.

Leave a Reply