Sentiment Analysis using text classification. A Naive Bayes Algorithm uses dictionaries of words to classify tweets as negatvie, neutral, or positive. Mood Correction submission are instantly applied. Retweets are black and white. Retweets appear in the Firehose for each instance of amplification. When captured like this, frozen in a snaphot, they look like duplicates. Each Retweet is considered a Tweet because of the Perspectival Agreement Intent of the Retweeter.