Wednesday, April 18, 2012

Data Science - Definitions

Found the following clarify the often heard "Data Science".

 1) What is data Science report from O'Reilly
     According to the author, Mike Loukidas, Data Science enables not just using web data, but creating data products. The author explains this with examples from the way Google uses the PageRank algorithm for search,  how corrections to misspelled searches are suggested, voice search integration to the core search engine, tracking the Swine flu epidemic of 2009, iTunes concept where music is viewed as data in the CDDB database, Amazon's "recommendation" systems, Facebook and LinkedIn relationship patterns are all data products. The second attribute is that users are in the feedback loop for these data products/ application systems. The author explains how data science is not traditional statistics and is to be seen as a holistic approach to data. 
   


       The course is taught by Jeff Hammerbacher (Facebook, Cloudera) and Mike Franklin. 
       Data preparation, presentation, products, observation and experimentation are the five 
      components of the course. The slides are available for public view.