Abstract
Social Networking Sites (SNS), such as Facebook and Twitter, are important tools for political campaigns. A line of related work analyzed political campaigns online. The initial efforts in analyzing campaign discourse functions relied on human analysis, which is time consuming and does not scale well with big data. To address these gaps, we propose a model to detect the type of campaign topics: Policy vs. Character, and how the public (commentators) responded to these messages. The proposed model yielded an accuracy of 78% (F-measure) in detecting post type. Moreover, experimental results show the analysis of commentators linguistic and psychological characteristics.