Saatviga Sudhahar, Intelligent Systems Laboratory, University of Bristol
We present a new computational methodology for large scale narrative analysis of news content, with an application to political discourse concerning US presidential elections. News articles concerning the elections are recognised, parsed, and used to generate a list of political actors and a network representing their political relations. That network is in turn analysed to extract information about the role that each actors play in the political narrative. The method is entirely automated and very scalable, and its results can be accessed via the project website. So far our system has analysed 125254 articles in 719 US and international news outlets extracting 31476 actors.