In this project, you write a program that takes the name of a recent movie, searches twitter for all tweets containing the name of the movie, figures out which tweets are actually expressing an opinion about the movie (and discards the rest), and then figures out the percentage of the positive reviews vs. negative reviews.
In other words, an algorithm that can automatically tell you twitter's rating of a movie. The output would be something like what you see here: Bol Bachchan Twitter Verdict - except that page is done by an actual human, and you need to do this automatically.
This is a classification problem - and the best algorithms for doing this are ones in which a training data set (consisting of manually classified tweets) is available at the time of development of the algorithm. This training data set will be provided to you based on data we've collected at http://wogma.com/