Rule based sentiment analysis system for analyzing tweets

In daily life, our beliefs and perceptions of reality are largely dependent on how other people analyze the world. People mostly share their opinions about different entities like organization, products, services etc. on social media, forums, and blogs. It's very difficult to analyze this enormous amount of data manually. Thus, sentiment analysis is the process which automatically identifies people's attitudes and emotional states. In this paper, sentiment analysis system has been proposed which computes the sentiment and score of each tweet extracted from Twitter about any parameter given by a user. The proposed system has achieved an accuracy of 86% in case of testing on 500 tweets and 94% while testing on 200 English sentences. The results given by the system are very stimulating and in future, it has potential to be used for the benefits of society.