Sentiment Analysis on Brazilian News Broadcast Data

Abstract

This work aims to compare the occurrence of negative emotion words in Brazilian broadcast news JN and JR, and also analyzes Twitter posts related to them. We use the Brazilian Portuguese version of LIWC dictionary, which is a Sentiment Analysis software. The results indicate that both JN and JR tend to use negative emotion words, but in JR this tendency is greater. Nevertheless, Twitter posts direct more criticisms towards JN than JR.