Presenting Sessions on Data Mining
I am currently preparing a few presentations on using Data Mining in business intelligence. These will be at the Brisbane SQL Server user group (this month) and SQL Rally Amsterdam (in Nov). I am especially looking forward to Amsterdam because this will be my first trip to the Netherlands.
The application of data mining within an organisation is an interesting topic for me which I liken to a milestone in organisational maturity. When most people discuss business intelligence (and analytics), they are talking about the production of a system so that the and end user can either get canned reports quickly or (more importantly) interrogate data in a real time manner. After all, this was the intent of OLAP! – the idea that the end user can drive data in real time so that once they think of a problem and a potential solution, they can verify their thoughts against data.
However, as good as this type of system is (compared to the starting base), this can be one of the short comings of stoping the BI journey here. That is, the user needs to think of a solution to a problem, and then verify against it. But what do we verify against? – and when do we stop investigating? This is one of the major benefits of data mining. It allows a scientific analysis at a targeted problem without the limitations of our dimensional thought (after all we can only think in a relatively small number of dimensions & attributes at a time).