Binary Separation Index for Echo Chamber Effect Measuring

In this paper, we develop a new metric to measure the echo chamber effect in online social networks for a particular topic. Unlike previous approaches, our one does not require the knowledge of users' opinions towards the fixed issue. Once information sources that spread relevant content in a social network are determined as well as their positions, our metric generates a value unambiguously. Changing from null to one, it increases as long as the level of informational separateness is on the rise. We discuss the problem of the metric calculation in the case of the social network Vkontakte. Considering attitude towards the Russian government as a topic, after empirical data processing we obtain the value of 0.802 that evidences the sufficiently high level of information separation among Vkontakte users.