This repo contains the java files for sentiment analysis of twitter data, the tweet corpus has also been attached with the "full" suffix file for creating the dictionary and the "test" suffix file for validation,user can enter the network configuration
stopwords.txt has also been uploaded for removing the words which do not contribute to a sentiment, you can update it by adding your own words or removing the already existing ones, its recommended to not to remove most of the words since the feature vector will increase
feature vector is prepared for all the tweets, then the training is done for the 80% tweets in the files twitter_positive for e.g. and validation is done on the twitter_positive_test for e.g.