Abstract—With the rapid development of e-commerce, financial industry and blockchain technology, cryptocurrencies have become a global phenomenon known to most people. The historical prices show that cryptocurrencies have experienced significant price fluctuations on both daily and long term valuations. Cryptocurrency market movement prediction systems have emerged to help people make an informed decision. Traditional supervised learning algorithms were used for predicting changes in cryptocurrency prices based on historical price data. Nowadays, the number of news available on the internet is increasing rapidly. Discerning the impact of news on price movement can provide a buying and selling advantage to an investor. In this paper, we describe a method for predicting cryptocurrency prices utilizing news and historical price data. Our paper analyses the ability of news data to predict price fluctuations for the second largest cryptocurrency in terms of market capitalization: Ethereum. The model is able to directly predict price direction by indicating whether to buy, sell, or hold. The final version of the model was able to correctly predict cryptocurrency price using historical data and sentimental information gained from news data. The important key in our model is the application of a set of natural language processing algorithms to identify the public moods for cryptocurrency fluctuations. We showed that sentiment analysis is an important perspective for cryptocurrency price prediction due to the interactive nature of financial activities.
Index Terms—Cryptocurrency market prediction, Ethereum, long short-term memory, machine learning, natural language processing, sentiment analysis, text mining.
The authors are with the Department of IT Convergence, University of Ulsan, Ulsan, 44610, Republic of Korea (e-mail: voanhdung@mail.ulsan.ac.kr, nqphuoc@mail.ulsan.ac.kr, okcy@ulsan.ac.kr).
[PDF]
Cite: Anh-Dung Vo, Quang-Phuoc Nguyen, and Cheol-Young Ock, "Sentiment Analysis of News for Effective Cryptocurrency Price Prediction," International Journal of Knowledge Engineering vol. 5, no. 2, pp. 47-52, 2019.
