With the rapid growth of the internet, millions of people are sharing their views and opinions on a variety of topics on micro blogging sites. On these websites user makes real time short and frequent posts about everything. These posts also include Sentiments which refers to emotions, feelings, attitude or opinion. Sentiment analysis is basically study of emotions and opinions from text. The basic idea is to analyze the results and predict outcomes that are based on customer feedback or opinions. It is helpful for consumers who want to find out the sentiment of products before purchase, or companies that want to monitor the public sentiment of their brands. Twitter sentiment analysis is tricky as compared to broad sentiment analysis because it contains slang words, misspellings and repeated characters. This research paper present the results of machine learning algorithms by classifying the sentiment of Twitter messages using distant supervision with the help of preprocessing steps needed in order to achieve high accuracy. The conclusion of this paper is presented by ten different sentiments from data taken.

 Keywords: Sentiments, Naive Bayes Classifier. Twitter, Machine learning algorithm.