Ethos, defined by Aristotle as the character of the speaker, plays a fundamental role in day to
day society being present in conversation to political debate. Despite this fact the automatic
extraction of ethos from natural language, ethos mining, has rarely been studied with the primary
focus in the argumentation community being upon the reasoning expressed in language (logos). In
this talk I will show the foundations of ethos mining, focussing upon the intricacies of UK
parliamentary debate and the natural language processing methods used in extraction. I will then
describe the current applications of ethos mining, in particular, the initial qualitative and
quantitative visualisations which can be expressed from ethos mining for individual politicians and
political groupings. Finally, I will explore possible future avenues of work in ethos mining,
especially improvements which can be made from a deeper understanding of the natural language and
the possibilities of further visualisation.
Rory is a PhD student within the Computational Ethos Lab (CELab), part of the Centre for
Argument Technology at the University of Dundee. His research has focussed on the creation of the
first corpus of manually annotated ethos and the first automatic extraction of ethos in political
debate, using Machine Learning and Deep Learning, as well as the visualisation of political
relationships and data trends. During his PhD, Rory has published at several international
conference venues most notably winning the best student paper award at the International Conference
on Computational Models of Argument (COMMA) 2016 and presenting at the International Joint
Conference on Artificial Intelligence (IJCAI) 2018.
University of Konstanz, Powerwall Room C202
The talk will be transmitted to Stuttgart, if technically possible.