Applying Voice Recognition to Vox Populi: State Transition Models in the Study of Public Opinion and Political Communication

21 Pages Posted: 13 Aug 2009 Last revised: 5 Oct 2009

Abe Gong

University of Michigan at Ann Arbor - Gerald R. Ford School of Public Policy

Date Written: 2009

Abstract

"Chunks" of related news coverage with the potential to affect political outcomes - are the building blocks of political narratives and an important concept in political communication. In this paper I make three important claims about news stories. First, consistent with common intuition, news tends to come in chunks. Second, this tendency can enrich studies of media and politics if used well, and lead to confusion and error if ignored. Finally, the methodological capacity to identify meaningful chunks of news on a large scale is within reach. I demonstrate how hidden Markov models can be used identify news stories within stream of media content, and discuss implications for the study of media, public opinion, and political communication.

Keywords: media, elections, public opinion, political communication, content analysis, hidden Markov models

Suggested Citation

Gong, Abe, Applying Voice Recognition to Vox Populi: State Transition Models in the Study of Public Opinion and Political Communication (2009). APSA 2009 Toronto Meeting Paper. Available at SSRN: https://ssrn.com/abstract=1450047

Abe Gong (Contact Author)

University of Michigan at Ann Arbor - Gerald R. Ford School of Public Policy ( email )

735 South State Street, Weill Hall
Ann Arbor, MI 48109
United States

Paper statistics

Downloads
33
Abstract Views
450