Automated construction and analysis of political networks via open government and media sources

Abstract

We present a tool to generate real world political networks from user provided lists of politicians and news sites. Additional output includes visualizations, interactive tools and maps that allow a user to better understand the politicians and their surrounding environments as portrayed by the media. As a case study, we construct a comprehensive list of current Texas politicians, select news sites that convey a spectrum of political viewpoints covering Texas politics, and examine the results. We propose a ”Combined” co-occurrence distance metric to better reflect the relationship between two entities. A topic modeling technique is also proposed as a novel, automated way of labeling communities that exist within a politician’s ”extended” network.