SQL Server data mining has always been somewhat of an overlooked, niche technology. There are several logical reasons for SQL Server data mining’s slow adoption. First, there are few companies that take the time to consider if they can benefit from the technology (OLAP tends to be the technological “stop point” in modern day). Furthermore, there are few Information Workers (IWs) and technology professionals that are trained in data mining concepts. Finally, SQL Server data mining requires the availability of an installed SQL Server Analysis Services instance.
I do believe that over time the first two reasons for SQL Server data mining’s slow adoption will dissipate. While the third reason is not going away, there is an alternative solution to the problem of SQL Server data mining requiring an installed instance of Analysis Services…The Cloud!
Azure, the Microsoft suite of cloud-based services includes a new technical preview of Data Mining in the Cloud “DMCloud”. DMCloud allows you to perform some basic data mining tasks leveraging a cloud-based Analysis Services connection. You can leverage a browser or download an Excel 2007 add-in to use DMCloud.
DMCloud is valuable capability for IWs that would like to begin considering SQL Server Data Mining without the added burden of needing a technology professional to first install Analysis Services. Additionally, IWs can use the DMCloud services no matter where they may physically be located as long as they have an Internet connection! The data mining tasks you can perform with DMCloud are the same Table Analysis Tools found in the traditional Excel Data Mining add-in. These data mining tasks include:
- Analyze Key Influencers
- Detect Categories
- Fill From Example
- Highlight Exceptions
- Scenario Analysis
- Prediction Calculator
- Shopping Basket Analysis
For more information on Microsoft’s Data Mining in the Cloud services, please see http://www.microsoft.com/azure/mining.mspx.