Discourse Analysis for Insights into Group Identity and Intent

ABSTRACT: For this project, we propose to modify our existing natural language processing (NLP) techniques to conduct automated discourse analysis on original documents produced by the People"s War Group (PWG) in India. Using templates developed by Toman et al (2010) we will evaluate our ability to automatically identify and measure shifts in in-group alliance building, out-group distancing, and the cognitive complexity of PWG senior leadership, as expressed in their writings in the group"s newsletter the People"s March. If successful, these techniques will facilitate the development of prototype software that will allow analysts to monitor, assess, and forecast, in near real time, the consequences of shifts in rhetoric employed by insurgent and extremist organizations. BENEFIT: Commercial applications will involve a data and analysis subscription service for clients in the defense, intelligence, and diplomatic domains, along with customized analysis and training.