Stance Classification using Dialogic Properties of Persuasion

Marilyn Walker,  Pranav Anand,  Rob Abbott,  Ricky Grant
UCSC


Abstract

Public debate functions as a forum for both expressing and forming opinions, an important aspect of public life. We present results for automatically classifying posts in online debate as to the position, or stance that the speaker takes on an issue, such as Pro or Con. We show that representing the dialogic structure of the debates, and agreement relations between speakers, greatly improves performance for stance classification, over models that operate on post content and parent-post context alone.