Events Details

Speaker:
Andreas Spitz (Associated Fellow / Data and Information Mining Lab)

Abstract:

A substantial majority of Americans share the belief that political discourse has become more negative and that the press is increasingly less reliable. However, as is often the case in politics, talk is cheap and hard data is difficult to come by. To obtain quantitative answers, neural language models allow us to exploit the parallelism in news reporting to extract and attribute politician's quotes from the news at web-scale. Using Quotebank, a comprehensive corpus of 235 million unique quotes from a decade of news, I will demonstrate how this data can be used to quantify trends in the use of political and journalistic language. In particular, I will focus on the uptick in negativity in U.S. politicians' language, quantify the shifts in language tone, and begin to unravel the decline in journalistic objectivity in the reporting of quotes in the news.