Big Data for Big Thinkers

October 31, 2011

“Big data analytics” is an emerging term in the storage industry that originated within the open source community to develop analytics processes that were faster and more scalable than traditional data warehousing.

Open source advocates hope to use data to extract value from the vast amounts of unstructured data produced daily by web users. I recently read an interesting Karmasphere write up called “Big Data IS Different— I Knew It!” in which Rich Guth mused about his past year spent at Karmasphere. In the period, his opinion of Big Data requires different analytic techniques than traditional business intelligence products provide. Guth asserted:

Today we announced version 1.5 of our Karmasphere Analyst product, a workspace for performing Big Data Analytics. It implements a new workflow for data analysts to mine and analyze Big Data.  We also released a whitepaper “Deriving Intelligence from Big Data in Hadoop – A Big Data Analytics Primer” that describes this workflow, discusses why this workflow is necessary and compares it to traditional BI and data warehousing approaches.

The challenge is to make clear exactly what “old methods” will not work and which “new methods” will work. As important, how does a person using a system with new Big Data methods determine if the outputs are accurate. Who wants to make a decision only to find out that the underlying set up of the new methods were off the mark. Most business intelligence professionals don’t know when an old and well worn method is delivering accurate outputs. Toss in a snappy graphic and the disconnect may become significant.

Jasmine Ashton, October 31, 2011

Sponsored by Pandia.com

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta