Abstract
Location based sentiment analysis is the use of natural language processing or machine learning algorithms to extract, identify, or characterize the sentiment content of a 'text unit', according to the location of origin of the text unit. In this paper, we study the application of location based sentiment analysis using Twitter for identifying trends and patterns towards the Indian general elections 2014. We perform data (text) mining on 650,000 tweets collected over a period of 5 days pertaining to two political parties in India, during the campaigning period. We make use of Naive Bayes algorithm to build our classifier and classify the test data (as positive or negative) according to it. We identify the sentiment of Twitter users towards each of the two Indian political parties, by location and plot our findings on an Indian map. In the end, we present our observations and conclusions and how certain "social events" influence the sentiments of Twitter users on the social network. We also discuss the issues related to geo-location using the data obtained from the Twitter API.