Mining Value from Massive Data Sources

Dr. Shah, team lead at Thomson Reuters R&D, discusses strategies that have (and have not) worked in dealing with massive datasets.

She draws from two projects at Thomson Reuters to create novel insights from large scale data:
1. Using automation to identify expert stock recommenders on Twitter.
2. "Language Magnet", a project that uses natural language processing to mine abnormalities in SEC filings.

01:05:12

This talk was filmed at the Machine Learning meetup at Pivotal Labs in New York.

 

Interested in Machine Learning? Check out our 3-day course Practical Machine Learning for Engineers, Nov 10-12th in NYC.