Sentiment Analysis in Financial Social Media: Impact on Market Volatility

This dataset accompanies the master's research "Sentiment Analysis in Social Media to Understand the Impact on Stock Market Volatility". It contains 953,209 Stocktwits messages collected between December 2024 and March 2025, classified by the FinTwitBERT-sentiment model. The data integrate social media sentiment indicators with market return data for multiple financial assets (NVDA, PLTR, NIO, LCID, EURUSD, CL=F, IRX), allowing econometric and time-series analysis of investor sentiment and market behavior.