Finding Intermediary Topics Between People of Opposing Views: A Case Study

Abstract

In micro-blogging platforms, people can connect with others and have conversations on a wide variety of topics. However, because of homophily and selective exposure, users tend to connect with like-minded people and only read agreeable information. Motivated by this scenario, in this paper we study the diversity of intermediary topics, which are latent topics estimated from user generated content. These topics can be used as features in recommender systems aimed at introducing people of diverse political viewpoints. We conducted a case study on Twitter, considering the debate about a sensitive issue in Chile, where we quantified homophilic behavior in terms of political discussion and then we evaluated the diversity of intermediary topics in terms of political stances of users.