SENTIMENT ANALYSIS ON TWITTER

Given a set of tweets pre-classified as positive or negative sentiment, I analyzed the word frequency patterns to develop a sentiment classification algorithm. Using this algorithm, I then classified another set of tweets to determine their sentiment. The data set I used was from a project using real tweets in 2009 (Go, A., Bhayani, R. and Huang, L., 2009. Twitter sentiment classification using distant supervision. CS224N Project Report, Stanford, 1(2009), p.12.). The original data was retrieved from Kaggle at https://www.kaggle.com/kazanova/sentiment140.