DiSCourse Seminar with Kohei Watanabe
21 February 2020, 12:00 (CET)
Digital Science Center, Ursulinenpassage, Innrain 15, Open Space Area (1 st floor)
DiSCourse - The Digital Science Seminar Series on
Measuring Geopolitical Threat over 150 Years through the New York Times: Technology and Methodology for Analysis of Ultra-sparse Data
The application of natural language processing techniques, called quantitative text analysis, is becoming increasingly popular among social scientists. In recent years, political scientists have been attempting to apply this methodology to capture geopolitical instability through the analysis of newspaper articles or foreign policy documents. In this seminar, Kohei Watanabe will present how his team successfully measured geopolitical threat for the United States over the last 150 years using the New York Times API.
The geopolitical threat index (GTI) not only contributes to the development of international political research by offering empirical data for America’s perception of its security, but embodies technological and methodological advancement in the field. With his project as an example, Watanabe will explain how researchers can efficiently analyze ultra-sparse textual data on an open-source software infrastructure that exploits highly skewed frequency distribution of words.
Kohei Watanabe, University of Innsbruck, Department of Political Science and DiSC