Professional Microsoft® SQL Server® Analysis Services 2008 with MDX

16.5. Summary

In this chapter you learned about data mining, what it is used for, and what specific algorithms are available for use in SSAS 2008. The most important answer is to the question, "How does it help your business?" If you are now a step or two closer to that answer, you're doing great.

After understanding the data mining algorithms supported by SSAS 2008, you drilled down on two, developing step-by-step Microsoft Decision Trees and Microsoft Clustering models using data from a relational data source. You also learned about OLAP mining models, where you essentially built a model on top of a cube. With the OLAP mining model you segmented the customers based on Internet Sales, and you were able to utilize the results of that mining model within a cube.

Aside from those off-the-shelf algorithms, SSAS 2008 provides a way to plug in your own data mining algorithm and/or data visualization capability (viewer). For details on this, please refer to the SSAS 2008 product documentation. In a nutshell, you utilize the plug-in architecture provided by SSAS 2008 and implement certain interfaces so that the server can utilize the results coming out of the algorithm you created. Once you have implemented your algorithm you can expose it using the SSAS server properties. For an in-depth understanding of data mining in Analysis Services 2008, we recommend you read Data Mining with Microsoft SQL Server 2008, by Jamie MacLennan, ZhaoHui Tang, and Bogdan Crivat.

SSAS aims at providing ...