CANDIDATE SENTIMENT SCORE
CampaignPop a sample set of approximately 24,000 tweets about the individual candidates each day (about 744,000 over the course of the month). Tweets are then analyzed using the Natural Language Toolkit (NLTK) to determine if their content is either mostly negative or mostly positive. Tweets that mention both candidates are ignored since the target of the negative or positive signals is difficult to determine. Tweets deemed as 'neutral' are also thrown out. The Sentiment Score is then calculated according to the following equation: (Total Positive Tweets)/(Total Positive Tweets + Total Negative Tweets) * 100. A perfect score of 100 is achieved if all of the tweets sampled are deemed positive.