Applying textmining to classify news about supply and demand in the coffee market

Abstract

This work verifies the feasibility of text classification using supervised machine learning method to promote the web news monitoring on factors that impact supply and demand for the coffee market. To this end, a device was develop that enables the empirical evaluation of the Naive Bayes method to sort news collected from the web according to the categories: positive or negative to supply and to demand. The tests show the feasibility of Naive Bayes classifier to identify factors that affect supply and demand in coffee market.