Mining Personal Media Thresholds for Opinion Dynamics and Social Influence

摘要

To study the detailed effects of social media consumption on personal opinion dynamics, we gather self reported survey data on the volume of different media types an individual must consume before forming or changing their opinion on a subject. We then use frequent pattern mining to analyze the data for common groupings of responses with respect to various media types, sources, and contexts. We show that in general individuals tend to perceive their behavior to be consistent across many variations in these parameters, while further detail shows various common parameter groupings that indicate response changes as well as small groups of individuals that tend to be consistently more easily swayed than the average participant.