CORPS: A Corpus of Tagged Political Speeches for Persuasive Communication Processing

In political speech, even if the audience is sympathetic to the speaker and does not need to be persuaded, it tends to react or respond to signals of persuasive communication (including an expected theme, a name, an expression, and the tone of the voice). In this article, we describe the creation of a corpus of political speeches tagged with audience reactions, such as applause, as indicators of persuasive expressions. We hypothesize that corpora of this kind can be usefully employed in the qualitative analysis of political communication. In addition, we present a corpus-based approach for persuasive expression mining that relies on techniques from natural language processing (NLP). We show how the approach can support the analysis of political communication, providing insights well beyond those of traditional word-counting analysis techniques.