Modeling of political web pages spreading in WeChat networks

Abstract

Modern social media has greatly facilitated the ability and efficiency of people to access and consume information, as well as intentionally or unintentionally spread political rumors and nationalist sentiments. This paper addresses the problem of modeling of political web pages spreading in WeChat networks. At first, a large number of web pages diffused in WeChat are collected, in which more than two hundred million users are involved. The widely disseminated pages are extracted and divided into two categories: political and non-political pages. Then the topological and temporal features of these web pages are analyzed and compared with respect to cascade size, life span, width, height, average depth, and average path length. The properties of involved user's behaviors are examined in terms of viewing delay, sharing delay, and sharing probability. At last, the Unknown-View-Share-Removed (UVSR) model is employed to characterize the dynamic diffusion process of political web pages. The model is driven and validated by the empirical observations of political web pages diffused in WeChat networks. Our findings contribute to predicting and even regulating political rumors and nationalist sentiments.