An Intelligent Interface for Organizing Online Opinions on Controversial Topics

Abstract

An enormous amount of posts and comments are shared in online social forums, which often organize these online social opinions based on semantic contents. However, for controversial topics, people with different attitudes and stances often have very distinct perspectives, reactions, and emotions to the same post. Organization by semantic contents often encourages selective exposure to information, which may exacerbate opinion polarization. To address this problem, we design a novel interface that allows people to better understand and appreciate people with different stances in social forums. Our interface was developed to allow interactive visualization and categorization of original posts about a controversial topic with crowd workers' reactions and emotions from different stances. We evaluated the interface using Reddit posts about US presidential candidates. Results demonstrate that the interface can mitigate selective exposure and help users to adopt a broader spectrum of opinions than the traditional Reddit interface.