Congressional Communication Networks on Various Social Issues

• In Event: Network Analysis

Abstract

Congressional communication has been an indispensable component of political
deliberation in the United States. Members of the congress communicate with one another on a wide array of social issues. The current study aims to address research questions by integrating topic modeling and network analysis techniques. The data of the study contains 366,093 tweets posted by the members of the 114th congress of the United States on Twitter from January 1, 2016, to August 31, 2016. The tweets are classified into 50 topics with Latent Dirichlet Allocation (LDA) topic modeling approach. Then congressional communication networks on these topics are constructed based on members’ retweeting and mentioning behavior on Twitter. We compared basic structural properties of each topic networks and examined their connections using Exponential Random Graph Modeling method. We believe that our findings provide new insights into the dynamics of legislator communications.

Authors

• Arram Bae, Michigan State U

• Winson Peng, Michigan State U