Geo-spatial Clustering of Sentiments on Social Media

Abstract

Social networking sites have tremendously captured online communication over the social web. With the growth in number of users on social networks, the social data has also grown exponentially. One of the predominantly used social networking sites includes Twitter. It is one of the most authenticate social platform that allows users to express their views on current trends and topics. Sentimental analysis of such dynamically changing user behavior upholds huge amount of contextual information. The behavioral data could be further evaluated to find the associated sentiments. Our research is focused on pre-processed analysis and classification of real-time tweets, based on the emotional content. Our novel approach applies density-based clustering with longitudinal locations from the tweets to reveal social communities for sentimental analysis.