ArguAna for the Web


Argumentation Analysis (ArguAna) for the Web is a cooperative research project of the Webis group at the Bauhaus-Universität Weimar and the UKP lab at the TU Darmstadt, funded by the German Research Foundation (DFG). The project targets at the specific challenges of mining arguments and their relations from natural language text on the web. We seek to establish foundations of algorithms that (1) robustly apply to various forms of web argumentation, (2) efficiently leverage the scale of the web, and (3) complement argument mining with an argumentation analysis to effectively assess important quality dimensions.

The rationale of the project is that people compare arguments in many situations, e.g., when buying products or when forming opinions on political controversies. The richest and most up-to-date argument source is the web. Previous research on argument mining tackles the identification and relation of arguments within a particular domain, but it does not suffice to successfully mine arguments from the web. As part of our research, we have developed a service for argument search (along with a REST interface), demonstration prototype for assessing the argumentation quality of persuasive essays, and provide demo access to the winning submission of our human value detection shared task. [api] [demos: essay scoring, human value detection] [service]


Students: Janek Bevendorff, Jonas Dorsch, Viorel Morari, Jana Puschmann, Jiani Qu, Patrick Saad