Polarization in social media

Abstract

The development of social media has changed the way in which information is consumed by the public. However, it also promotes polarization, especially among the more controversial topics. Furthermore, recommendation systems commonly used in social media have been shown to emerge an echo chamber effect, amplifying user groups’ polarization. Most of current studies focus on analyzing the generation of polarization phenomena on social media but rarely investigate how to quantify polarization. In this study, we construct user opinion networks and utilize random walk to quantify the polarization of related topics. The experiments on three real datasets, i.e., Bilibili, YouTube and Reddit, demonstrate that there is polarization in Bilibili and YouTube, especially in Bilibili. Our work complements quantitative measurements of polarization in social media platforms.