International Journal of Computational Linguistics Research Durgesh M. Sharma, Mirza Moiz Baig 6 2 2015 Sentiment analysis is a process to determine the view on targeted keyword from the internet which is unknown for the users. Users are the customers who share their sentiments on Social Network sites and it make a valuable platform for tracking and analyzing the public sentiments. Tracking and analyzing the sentiments need to extract the sentiments from the internet and there is one site i.e. Twitter. It is a micro blogging site that allows people to share their opinions in 140 characters only. So, mining can easily be done. In this work, we use Twitter for extracting tweets using TF*IDF algorithm and without using any special software we would apply the labeling technique for finding the polarity on tweets viz. positive, negative or neutral.