Estimating Subjective Argument Quality Aspects From Social Signals in Argumentative Dialogue Systems

Information about a subjective user opinion towards an argument is crucial for argumentative systems in order to present appropriate content and adapt their behaviour to the individual user. However, requesting explicit feedback regarding the discussed arguments is often impractical and can hinder the interaction. To address this issue, we investigate the automatic recognition of user opinions towards arguments that are presented by means of a virtual avatar from social signals. We focus on two different user opinion categories (
convincing
and
interesting
) and two different types of social signals (facial expressions and eye movement). The recognition is addressed as a supervised learning problem and realized using the argument search evaluation data discussed in previous work. The overall performance is compared to a human annotation on a subset of the collected data. The results show that the machine learning performance is similar to human performance in both recognition tasks.