Sentiment Analysis for Twitter

Jordan Marah

Supervisor: Greg Doyle

This project was undertaken for my final year project for 4th year Software Development in the Institute of Technology, Carlow. The purpose of this project was to attempt to try and derive sentiment from a tweet. I chose this project as I've been a Twitter user for almost a decade and thought it would be interesting to try and train a machine-learning model that was capable of classifying tweets as either positibe or negative. The technologies I used in this project were Scikit-learn and Keras for training the Support Vector Machine and I used Flask with SQLite and SQLAlchemy for the web application, Tweepy is the the Twitter API used to retrieve tweets. The application allows users to search specific terms, words, or topics and the application will return the sentiment of the given tweets that contain the query.