Social media platforms such as Twitter or Reddit have become an integral part in political opinion formation and discussions, accompanied by potential echo chamber forming. In this paper, we examine the relationships between the interaction patterns, the opinion polarity, and the socio-demographic characteristics in discussion communities on Reddit. On a dataset of over 2 million posts coming from over 20k users, we combine network community detection algorithms, reliable stance polarity annotations, and NLPbased socio-demographic estimations, to identify echo chambers and understand their properties at scale. We show that the separability of the interaction communities is more strongly correlated to the relative socio-demographic divide, rather than the stance polarity gap size. We further demonstrate that the sociodemographic classifiers have a strong topical bias and should be used with caution, merely for the relative community difference comparisons within a topic, rather than for any absolute labeling.

No abstract