Twitter Sentiment Analyzer (P-0631) (MFN 6779)

Abstract

Twitter Sentiment Analyzer is a sentence based sentiment analyzer. This project aims to analyze the orientation of "Tweets" over Twitter ofthe people of Pakistan and give an overview of the political orientation of Pakistanis over a period of events. Users will be able to search about a party using hashtags, the respective data will then be taken from our data dumps and/or taken from Twitter. Our sentiment classifier, trained over the algorithm of Naive Bayes Classifier, will divide the tweets into positive, negative and neutral tweets. The results will then be displayed over a dashboard for easy understanding. Our vision for this project was to create a one click solution to find out about the political orientation of the People of Pakistan. This will help us estimate where the political tides would take us in electoral events. This also helps political analysts by gauging what kind of route a party may take. This also provides insight to political parties as they would gain a clearer overview of how happy or displeased people are about their policies or about their stances taken in political events. Our application targets Twitter users tweeting in English. This decision was done in order to include the maximum number of people as possible in our analysis. The final version is free of cost and open to the general public.