Tweet Sentiment Classifier is a program built in Python that uses Data Mining concepts like Naive Bayes Classification to classify the tweets into positive, negative, neutral or mixed sentiments. The input data is the 2012 presidential election (between Obama and Romney) tweets. This training set consists of about 7000 tweets for each Obama and Romney and the test set consists of about 2500 tweets. The program learns from the training set and applies the classification on the test set and generates the Accuracy, Precision, Recall, F-Score and Confusion Matrix.