Abstract

Individual’s motivations for participating in social networks can vary greatly, and this heterogeneity is reflected in the wide-variety of behaviors and communication styles we see in large social networks. In this paper, we present a first study of the basic properties of communication in a heterogeneous-intent network and their implications. First, through both user studies and large-scale data analyses, we validate that individual’s specific intents affect their style of communication and their selection of messages. Secondly, we propose a simple structural model of information diffusion within a heterogeneous-intent network and present the results of experiments exploring its emergent properties, finding that it recreates aspects of empirically observed cascades that are not captured in previous models.