Bitcoin Price-Predictive Model Based on Twitter News

Bitcoin is a volatile investment and prices are hard to predict, with investors losses estimated at 70% of their portfolio from the market peak of Nov. 2021. This volatility in the price presents a great opportunity for investors. While there are various techniques to predict the Bitcoin price, sentiment analysis ranks highest in terms of prediction accuracy. Specifically, Twitter sentiment has shown to be valuable in predicting the increase or decrease in Bitcoin’s price.This praxis uses news tweeted on the Twitter social media platform to predict the direction of Bitcoin price for the next day. The model developed aims at improving the accuracy of the prediction of the direction of Bitcoin price by implementing advanced machine learning techniques as well as state of the art algorithms for sentiment analysis. In addition, to the sentiments, the model also incorporates reaction to the news by factoring in the number of likes and retweets to further improve the prediction accuracy. The results obtained not only emphasize the importance of news in predicting the direction of Bitcoin price, but also, sets a new benchmark for predicting the direction of Bitcoin next day price achieving a 65.2% accuracy of prediction. The solution presented has many practical and useful applications and can be expanded to other areas that can benefit from the application of social media sentiment analysis to predict future asset prices (ex. real estate, stocks, etc.)

Author

• El-Youssef, Rami

Language

• en

Keyword

• Bitcoin
• Predictive Model
• Sentiment Analysis
• Machine Learning

Date created

• 2022

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Engineering Management

Degree

• D.Engr.

Advisor

• Sarkani, Shahryar
• Fossaceca, John

Committee Member(s)

• Fossaceca, John
• Blackford, Joseph P
• Sarkani, Shahryar

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/d217qq31k