At the COMPUTATIONAL RHETORIC WORKSHOP Computing Figures, Figuring Computers II

I'm in Waterloo, ON, at the University of Waterloo, at "Computing Figures, Figuring Computers II" -- the second workshop on computational rhetoric.

The lecture I'm listening to right now is on "Argument Mining," a presentation by faculty from the Center for Argument Technology at the University of Dundee.  (Previous publications from these faculty can be found here: "Combining Argument Mining Techniques.")

Four types of argument mining techniques are applied to a corpus of written and spoken texts:  In addition to the three previously published about...

...the scholars are looking at mining for rhetorical figures.

This work is deeply linguistic.  At the basic level, they look for signal words...

But in a way that fascinates me, they can code machinery to look for argument structures like "appeal to expert opinion" and "anaphora."

This is cutting edge stuff, systematizing the catalogs of tropes, schemes, figures that characterized Old Rhetoric textbooks.