Skip to main content

Beyond Aggregations: Other Ways to Analyze Data


About This Video

Aggregations have been the pathway from data to knowledge: reporting, OLAP, PowerPivot,… But all these techniques basically compute sums (or other aggregates) over large amounts of data. In this talk we demonstrate that important information can be overlooked if we only apply aggregations in our data analysis. Sequences of errors that often repeat, clusters of customers with similar buying patterns, summerizing a stream of Twitter messages, it are all examples of tasks which aggregates can not easily handle. But don’t panic: Analysis Services can analyse your data in another way.

In this demo-driven session we play around with different data sources, including SQL Server logfiles (yes, this talk is also useful to administrators) and social media feeds (yes, this talk is also useful to social people), and see whether mining models can tell us more that plain aggregates do. Along the way, we have a brief look at how to build models, and how front-ends can use them.

SQL Server Days 2011 is brought to you by the Belgian SQL Server User Group ( VZW) in collaboration with Microsoft Belux.


Published Date: January 16, 2012
Presented By: Nico Jacobs




Video: WMV | MP4