Opinion Mining and Sentiment Analysis on Twitter

In today’s world of Internet, people became more eager to express and share their opinions on web regarding day-to-day activities and global issues as well. In this world of Internet various communication forms have emerged, by using these communication forms people can express their ideas, opinions about things which are going around them. Twitter is one those communication forms which are getting popularity now days among the people for interacting and expressing opinion in online word. Tweets on twitter provides an idea about peoples reaction toward particular event, product etc. This paper describes how sentiment analysis is done using tweets with text and emoticons. Tweets are collected for five different topics for implementing topic level search instead of the keyword based by using LDA algorithm. For sentiment analysis Bayesian logistic regression algorithm is used. By using twitter API tweets will be collected and after preprocessing the collected tweets classifier is applied to find the peoples sentiment towards given topic are positive, negative, strongly positive, strongly negative, or neutral.