|Project Description: ||Social media is the commonplace that allows individuals to create an account, share their views or opinions. Some of the social media are Facebook, Twitter, Instagram, etc. Twitter is one of the social media that allows the user to share the ideas within the character count of 140. Twitter works as an excellent platform for sentiment analysis. Anyone can follow anyone to receive the tweets to track the person in their timeline. As a matter of fact, the tweets of any account is visible to anyone who has a Twitter account. As a result, sentiment analysis has become challenging research area in the past decade.
In the proposed system, we use Twitter 4j interface to connect our application with Twitter. The sentiment of the tweets is analyzed based upon the Naive Bayes classifier. After analyzing the sentiment of the tweet, we connect and mine the Twitter data and store them in a MySQL database. Naïve Bayes classifier is used on mined data to get positive and negative tweets. These label tweets are taken to analyze and further processed to sentiment normalization. Finally, the result of Naïve Bayes and sentiment normalization are combined to get accurate results. This sentiment analysis is performed for secured and non-secured data that are collected from Twitter. The proposed system can be used for alerting the users in advance before any disaster or event is going to happen in a public place.