Visualizing the Extension Space of Abstract Argumentation Frameworks Using Clustering Methods

From International Center for Computational Logic
Toggle side column

Visualizing the Extension Space of Abstract Argumentation Frameworks Using Clustering Methods

Master's thesis by Mei Yang
Abstract argumentation is a very active research area in the field of argumentation. Currently, its visualization methods focus on displaying the details of every single argumentation extension in isolation. In this thesis, we present a new visualization approach, which clusters argumentation extensions based on how much they overlap and provides an overview of the whole space of extensions. Our implementation, called Neva, visualizes not only the extensions in low-dimensional spaces but also the argument corre- lation and argument distribution. Our application shows that clustering on argumentation extensions is feasible.