Price Movement Prediction of Cryptocurrencies Using Sentiment Analysis and Machine Learning

TLDR

Cryptocurrencies are an emerging market with low entry barriers and abundant data, making them ideal for studying market behavior via sentiment analysis and machine learning, though prior work has largely focused on Bitcoin. The study aims to predict price movements of Bitcoin, Ethereum, Ripple, and Litecoin using common machine learning tools and social media data. We compare neural networks, support vector machines, and random forests using Twitter and market data features to predict cryptocurrency price movements. Results demonstrate that machine learning and sentiment analysis can predict cryptocurrency markets, with Twitter data alone sufficient for some currencies and neural networks outperforming SVM and random forest models.

Abstract

Cryptocurrencies are becoming increasingly relevant in the financial world and can be considered as an emerging market. The low barrier of entry and high data availability of the cryptocurrency market makes it an excellent subject of study, from which it is possible to derive insights into the behavior of markets through the application of sentiment analysis and machine learning techniques for the challenging task of stock market prediction. While there have been some previous studies, most of them have focused exclusively on the behavior of Bitcoin. In this paper, we propose the usage of common machine learning tools and available social media data for predicting the price movement of the Bitcoin, Ethereum, Ripple and Litecoin cryptocurrency market movements. We compare the utilization of neural networks (NN), support vector machines (SVM) and random forest (RF) while using elements from Twitter and market data as input features. The results show that it is possible to predict cryptocurrency markets using machine learning and sentiment analysis, where Twitter data by itself could be used to predict certain cryptocurrencies and that NN outperform the other models.