The unprecedented adoption of social media for communicating political views has created widespread opportunities to study the opinions of enormous numbers of people who have been labeled as politically active individuals in real time. But the absence of methods to classify between users of conflicting po- litical alignments, absence of distinct signals at the very base level i.e. for each individual may, in the aggregate, mystify extreme partisan differences in ideologies that are important and associated with a particular political strategy. In this paper, we present a preliminary yet ground-breaking study of the social representation of news articles on Twitter. Existing methods for predicting the political inclination of users use techniques which rely on hand- engineering and use a small number of candidate predictors based on domain knowledge. Moreover, literature consists of comparative analysis of extreme ideology with neutral, small scale personality analysis and no mapping between network analysis and political inclination. In this work, we perform diverse feature analysis, a comparative linguistic analysis of the aggregate and top left and right-biased user base, study the personality traits on a large scale using IBM Watson Personality Insights API, perform network analysis using the follower-followee & reply graph & then build a set of machine learning models that are able to automatically detect the bias associated with a tweet.