by Angela Guess
Forbes contributor Meta S. Brown recently wrote, “Many IT pros are sincerely interested in supporting analytics, but don’t know how to differentiate legitimate business needs from nice-to-haves, and are frustrated that data analysts don’t understand and appreciate the reasons behind IT practices. In “=What Big Data Analytics Professionals Want From IT I explained the most fundamental things that analysts need to do their job, things like raw (not aggregated) data and adequate computing power. Now, I’ll take this theme a little deeper to show you the major elements of a good analytics process, and explain some ways that IT can and should be involved.”
Brown continues, “The Cross-Industry Standard Process for Data Mining, better known as CRISP-DM, has been around for more than a decade, and it’s by far the most widely-used analytics process standard. It’s an open standard, which anyone may use, developed by a consortium of over 200 interested organizations, with funding from the European Union. While it was developed specifically for data mining, it is flexible enough to suit many analytic styles. This process model is now so popular that diagrams from the original CRISP-DM documentation often crop up in presentations, without any reference to the original source of the material, or even to data mining.”
She goes on, “The CRISP-DM process model has six major phases: (1) Business Understanding: Get a clear understanding of the problem you’re out to solve, how it impacts your organization, and your goals for addressing it. (2) Data Understanding: Inspect, describe and evaluate the available data. (3) Data Preparation: Take data from the state it’s in to the state needed for analysis. (4) Modeling: Use mathematical techniques to make models (equations or other logic) you can use to support business decisions. (5) Evaluation: Figure out whether your models are any good. (6) Deployment: Integrate models into everyday business.”
photo credit: Flickr/ bluefountainmedia
