Este capítulo foi apresentado no evento Brazilian Symposium on Multimedia and the Web (WebMedia), 2019

Abstract: The World Wide Web (WWW) is one of the main sources of opinions nowadays. Whether through news sites, blogs or social networks, a lot of information is available to explore. Thus, a great amount of applications has arisen to explore this knowledge through automatic sentiment and opinion...

Abstract: The World Wide Web (WWW) is one of the main sources of opinions nowadays. Whether through news sites, blogs or social networks, a lot of information is available to explore. Thus, a great amount of applications has arisen to explore this knowledge through automatic sentiment and opinion extraction. The technique known as Sentiment Analysis aims to analyze opinions, sentiments and emotions present in unstructured data. Among its applications, many papers have been addressing the impact of news and social media publications on the financial market. However, the literature lacks comparisons of the impact that different online information sources have on the financial market. Therefore, this work analyzes the impact of sentiment from news and tweets, written in Portuguese, on the movement of the Brazilian stock market (evaluated through the Ibovespa index). Data from 01-Jan-2019 to 31-Mar-2019 were considered and the Sentiment Analysis was made with an Artificial Neural Network architecture known as Multilayer Perceptron (MLP). The results indicate that the sentiment from news can be more accurate to forecast the movement of the financial market when compared to tweets, both with respect to the opening prices and transaction volume

Fechado