Traditional social forums often organize social opinions about controversial topics based on semantic features. This may encourage selective exposure, which could cause social polarization. In our project, we designed and implemented an interactive interface which allows users to visualize and categorize social opinions based on people's emotional reactions. We found our interface could mitigate selective exposure through evaluation.
Currently, to organize social opinions, people focus on the semantic features of the text in the post and the comments. However, in reality, people may not express their reactions of a post and their stances in the text of the comments they made.
To satisfy people’s demand to see others’ reactions and stances, we designed an interface to show people’s different stances and emotional reactions in different granularities of each topic. In addition, we will also show the degree of controversy of each topic and provide users with a social opinion filter to select post which caused diverse stance and reaction from people. Our work could be treated as a complimentary work to the current solution which only focus on semantic features.