2026-05-24T00:39:33Zhttps://keep.lib.asu.edu/oai/request

oai:keep.lib.asu.edu:node-1315742024-12-18T20:58:30Zoai_pmh:alloai_pmh:repo_items

131574 https://hdl.handle.net/2286/R.I.56421 http://rightsstatements.org/vocab/InC/1.0/ 2020-05 19 pages eng Steinberg, Sam Boscovic, Dragan Davulcu, Hasan Computer Science and Engineering Program Barrett, The Honors College In this paper I defend the argument that public reaction to news headlines correlates with the short-term price direction of Bitcoin. I collected a month's worth of Bitcoin data consisting of news headlines, tweets, and the price of the cryptocurrency. I fed this data into a Long Short-Term Memory Neural Network and built a model that predicted Bitcoin price for a new timeframe. The model correctly predicted 75% of test set price trends on 3.25 hour time intervals. This is higher than the 53.57% accuracy tested with a Bitcoin price model without sentiment data. I concluded public reaction to Bitcoin news headlines has an effect on the short-term price direction of the cryptocurrency. Investors can use my model to help them in their decision-making process when making short-term Bitcoin investment decisions. Bitcoin Long Short-Term Memory Neural Network Sentiment Analysis Web Mining Twitter Predicting Bitcoin Price Trend using Sentiment Analysis