PREDICTING THE VALUES OF CRYPTOCURRENCIES USING HISTORICAL VALUES, BLOCKCHAIN INFORMATION AND TWITTER SENTIMENT
DOI:
https://doi.org/10.24867/04BE13MilutinovicKeywords:
sentiment analysis, prediction, BitcoinAbstract
This paper studies the problem of Bitcoin price prediction based on historical prices and blockchain information. As the price of a cryptocurrency is dictated largely by speculation we also examine the influence of twitter sentiment on the performance of our predictive models. We have also experimented with historical data for three other popular cryptocurrencies: Litecoin, Ethereum and Ripple. We experimented with three techniques for sentiment mining: convolutional neural networks, ensemble of lexicons and linguistic rules. The results of sentiment analysis were integrated with other data to train a recurrent neural network with GRU cell as our price prediction model. We evaluated our model both as a binary classifier (predicting whether the price will go up or down) and a regression predictor (predicting the actual price). The accuracy of the classification model was 57.3%, while the relative accuracy of the regression model was 99.39%.
References
[2] Blockchain. (2018, April 13). Retrieved April 13, 2018, from https://en.wikipedia.org/wiki/Blockchain
[3] Hutto, C.J. & Gilbert, E.E. (2014). VADER: A Parsimonious Rule-based Model for Sentiment Analysis of Social Media Text. Eighth International Conference on Weblogs and Social Media (ICWSM-14). Ann Arbor, MI, June 2014
[4] Augustyniak, L., Kajdanowicz, T., Szymanski, P., Tuliglowicz, W., Kazienko, P., Alhajj, R., & Szymanski, B.K. (2014). Simpler is better? Lexicon-based ensemble sentiment classification beats supervised methods. 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM 2014), 924-929.
[5] S.-H. Na, Y. Lee, S.-H. Nam, and J.-H. Lee, “Improving opinion retrieval based on query-specific sentiment lexicon,” in Advances in Information Retrieval, ser. Lecture Notes in Computer Science, vol. 5478. Springer Berlin / Heidelberg, 2009, pp. 734–738.
[6] K., & Y. (2014, September 03). Convolutional Neural Networks for Sentence Classification. Retrieved April 13, 2018, from https://arxiv.org/abs/1408.5882
[7] Wikipedia contributors. (2018, March 26). Bag-of-words model. In Wikipedia, The Free Encyclopedia. Retrieved 20:07, April 13, 2018, from https://en.wikipedia.org/w/index.php?title=Bag-of-words_model&oldid=832576977
[8] Liu, Y., Qin, Z., Li, P., & Wan, T. (2017). Stock Volatility Prediction Using Recurrent Neural Networks with Sentiment Analysis. Advances in Artificial Intelligence: From Theory to Practice Lecture Notes in Computer Science,192-201. Doi:10.1007/978-3-319-60042-0_22
[9] J. (2018, January 29). Jefferson-Henrique/GetOldTweets-python. Retrieved April 13, 2018, from https://github.com/Jefferson-Henrique/GetOldTweets-python
[10] Go, A., Bhayani, R., Huang, L.: Twitter sentiment classification using distant supervision. CS224N Project Report, Stanford (2009)
[11] Thelwall, M., Buckley, K., Paltoglou, G.: Sentiment strength detection for the social web. Journal of the American Society for Information Science and Technology 63(1), 163–173 (2012)
[12] SenticNet. (n.d.). Retrieved from https://sentic.net/
[13] A. L. Maas, R. E. Daly, P. T. Pham, D. Huang, A. Y. Ng, and C. Potts, “Learning word vectors for sentiment analysis,” in The 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies-Volume 1. Association for Computational Linguistics, 2011, pp. 142–150.
