Investigating the Relationship Between Twitter Sentiment and Bitcoin Price Movements

Abstract

Bitcoin, launched in 2009 as the pioneer cryptocurrency, has caught the eye of investors worldwide due to its nature and remarkable growth. Recent studies indicate that elements, like announcements, news sentiments, government regulations, and overall market sentiments significantly influence the trajectory of bitcoin price. This work presents a comprehensive analysis to provide deeper insights into the relationship between user sentiments expressed as tweets and bitcoin prices (daily price changes). As a significant contribution, the authors propose multiple approaches to performing the non-trivial task of integrating tweet sentiment with bitcoin price data. The chapter also presents the incorporation of inherent lag in the expression of sentiments and their impact on price change by demonstrating lagged sentiment analysis. The work employs state-of-the-art machine learning and deep learning models for analyzing the data both as classification and regression tasks to uncover hidden patterns.