Bitcoin movement prediction with text mining

Abstract

In the last few years, Bitcoin is one of the most discussed and popular topic in financial system. This article aims to predict Bitcoin movement by using Machine Learning and Text Mining models. Many models have been used to this end, including the most popular models in financial prediction; Artificial Neural Network (ANN), Support Vector Machine (SVM) and Logistic Regression (LR). In addition to this, in order to examine the effect of daily news on Bitcoin movement, the text mining models are involved into the prediction system. This paper focuses on applying Machine Learning models on a integrated dataset, which contains both historical Bitcoin values and features from daily news text. Overall, our model can estimate the direction of Bitcoin with a high success.