Prediction of cryptocurrency values using sentiment analysis of news and tweets

Authors

  • Wagner Resende Santos PUC Minas
  • Hugo Bastos de Paula PUC Minas

DOI:

https://doi.org/10.5335/rbca.v12i1.8831

Keywords:

Bitcoin, Cryptocurrencies, News, Prediction, Twitter

Abstract

The virtual currencies (or cryptocurrencies) are increasingly becoming more competitive in the global market,
atracting investors to earn profit on the oscillation of this market. These investiments are oriented by a simple
principle: buy the currencies when their market value is about to rise, and sell when their market value is
about to to drop. The goal of this work is to develop a model that predicts the movement of cryptocurrencies’
prices that is based on the public perception of people about the currencies. Predictions were made based
on news and tweets, and the prediction based on tweets achieved the best results, with 73% MDA using the
model created with XGBoost, being able to predict the results also during periods of oscilation. It is assumed
that the best results were obtained from tweets’ sentiments due to the bigger volume of data generated in a
single day, and due to the speed that public opinion manifests in the social network.

Downloads

Download data is not yet available.

Published

2020-03-23

Issue

Section

Original Paper

How to Cite

[1]
2020. Prediction of cryptocurrency values using sentiment analysis of news and tweets. Brazilian Journal of Applied Computing. 12, 1 (Mar. 2020), 1–15. DOI:https://doi.org/10.5335/rbca.v12i1.8831.