Stephen Baker at Business Week points to a post by Jeff Jonas on how companies are generating so much data so quickly that companies can’t make sense of it, leading to “enterprise-amnesia“. Both use this as a call for faster, smarter algorithms. Now far be it for me to be negative about such a cry (I like algorithms!) but I think they are missing an important step in the process: the modeling step. Modeling is formulating the objectives, constraints, and decisions in order for the algorithms to go to work. We’ve had lots of algorithms hanging around for decades that would address business issues, if only people would create appropriate models to use them. Not every problem needs a fancy new algorithm!
Jonas’ blog has a number of interesting posts on things like persistent analytics. This is all from a heavily information technology perspective, but OR should have a role to play in the issues he brings up.
Interesting thoughts. Without modeling the endless amounts of information go to waste. In the organisation i work in information is always construed to be positive to reflect positively on the department