Piero Ferrante will provide an introduction to text mining in addition to working through a functional application. Along the way he will also touch on key concepts behind web scraping, parallelization (mapreduce), and sentiment analysis using a host of open source solutions (namely Python, R, and RapidMiner). Although this particular example will focus on mining corporate earnings calls (presumably for integration into an automated or semi-automated trading strategy), the intent is to keep things general enough so that alternative applications become apparent.

Piero is the manager of analytics and insights at Blue Cross Blue Shield of Kansas City and longtime data enthusiast. He holds a BS in Finance and MIS from the University of Delaware and a MS in predictive analytics from Northwestern University.
Matt Habiger will show how one can use Amazon's Web Services (AWS) to handle large datasets. This can be particularly useful when the dataset is too large to fit into your personal machine's memory or you are running an algorithm that can be parrallelized.