Desktop Survival Guide
by Graham Williams
Many data mining projects work with ill defined and ambiguous business goals. Whilst the first reaction to such an observation is that we should be become better at defining the business problem, the reality is that often the problem to be solved is identified and refined as the data mining project progresses. That's natural.
The initiating meeting of a data mining project will often involve the data miner, the business expert and the data analyst. It is critical that all three come together to initiate the project. Their different understandings of the problem to be tackled all need to come together to deliver a common pathway for the data mining project. In particular, the data miner needs to understand the business perspective and understand what data is available that relates to the problem and how to get that data, and identify what data processing is required prior to modelling.