A Comprehensive Review of Social Media Data in Election Prediction

This project focuses on building a sentiment analysis system to aid in election prediction by analysing tweets related to political leaders. It leverages machine learning, specifically the Passive Aggressive Classifier, and incorporates text preprocessing techniques such as TF-IDF vectorization to classify tweets as either positive or negative. A user-friendly Flask web application is developed, allowing users to upload CSV files containing tweets. The application processes the data and displays the sentiment distribution through a pie chart visualization. By capturing public sentiment on social media, the system offers valuable insights that can contribute to forecasting election outcomes.