Many essential components of language charted by rhetoric, the ancient study of
persuasion, remain understudied and underrepresented in current Natural Language
systems. Our goal is to combine linguistic and rhetorical theories with discourse analysis
and machine learning to develop formal models of computational rhetoric that may be
usefully applied in real-world Computational Linguistics systems. As part of this initiative, we
are building an ontology of rhetorical figures and formalizing their expression.