Wall Street Bets Sentiment Analysis

Project Overview

This paper aims to study the correlation between Reddit's Wall Street Bets subreddit sentiment analysis and the performance of top S&P 500 companies. I employed natural language processing tools to analyze sentiment data.

Download paper View on GitHub

Key Achievements

• Built multiple linear regression models for each company and performed comparative analysis to evaluate the impact of 'wallstreetbets' sentiment across different companies
• Scraped financial data from Yahoo Finance and Reddit using APIs, collecting data for Apple, Microsoft, Google, Amazon, Meta, AMD, Nvidia Corporation, PepsiCo, Tesla, and Netflix from October 2021 to October 2022
• Cleaned and organized Reddit posts by month, and utilized VADER sentiment analysis tool to obtain compound sentiment scores, considering distinctive social media post qualities such as capitalization, exclamation marks, and emojis
• Employed Python modules such as pandas, sklearn, and APIs for efficient data extraction, data manipulation, and model implementation
• Observed that companies with higher post volumes tended to have better-performing models, indicating the relevance of sentiment analysis in stock market predictions

Tools and Technologies Used

• Python, Pandas, numpy, Scikit-learn
• Reddit API, Yahoo Finance API
• VADER Sentiment Analysis