README for sentiment.py Created by Scott Tarlow with help from Mika Wang and Adrian Hodge, 2016
sentiment.py uses a Naive Bayes classifier to predict the sentiment of a tweet. Included are an example training set and set of tweets Inputs: Training_Set.txt, _input.txt file Output: Predicted Sentiment (-1,0,1) of each tweet in a text file
This code can be easily modified for anyone's uses. The rating system for the sentiment is determined by the training set:
"string value" "sentiment score"
see Training_Set.txt for expected format of the training set.
This can determine the sentment of any string of words, given a large enough Training Set.
Raw output can be analyzed by any software you like!