Social media sentiment & stock prices
Can the language people use online help explain movements in financial markets? This research project explores that question through social media sentiment and stock-price data.
The question
TwitterSentiment brings together two very different kinds of information: the language of social media posts and numerical stock-price histories. The goal is to explore whether sentiment and market movements are related, rather than assume that one reliably predicts the other.
Inside the repository
The public repository contains Python notebooks, collected tweet and price datasets, a scraping component, and model experiments. Its dependencies include pandas, NumPy, Matplotlib, scikit-learn, and TensorFlow. The README describes LSTM and dropout experiments, along with unresolved data-shaping issues.
A research project in progress
This is exploratory research. The repository documents model limitations and open implementation questions, so its results should not be treated as a validated forecasting system or investment advice. The source provides a useful view into the process of joining text analysis with time-series modelling.