Rhetorical strategies in German argumentative dialogs

Abstract

Abstract. An important factor of argument mining in dialog or multilog data is the framing with which interlocutors put forth their arguments. By using rhetorical devices such as hedging or reference to the Common Ground, speakers relate themselves to their interlocutors, their arguments, and the ongoing discourse. Capitalizing on theoretical linguistic insights into the semantics and pragmatics of discourse particles in German, we propose a categorization of rhetorical information that is highly relevant in natural transcribed speech. In order to shed light on the rhetorical strategies of different interlocutors in large amounts of real mediation data, we use a method from Visual Analytics which allows for an exploration of the rhetorical patterns via an interactive visual interface. With this innovative combination of theoretical linguistics, argument mining and information visualization, we offer a novel way of analyzing framing strategies in large amounts of multi-party argumentative discourse in German.

VIANA: Visual Interactive Annotation of Argumentation

“…VIANA highlights keywords that are of specific interest to the annotation based on pre-defined word lists [25]. Connectors like "so", "if", "because" or "for" are bold and keywords like "appreciate", "promise" or "complain" that are associated with speech acts are italicized and underlined.…”

“…The system first enumerates so-called "elementary discourse units" [25]-(sub)sentences delimited by punctuation and clausal connectors-as fragments. Each fragment is represented as a tuple (embedding, label, weight, points, state).…”

“…To avoid a cold-start of the recommendation system, we initialize it with the output of a linguistic discourse-unit annotation pipeline [25] and refine it throughout the annotation process. The pipeline identifies discourse units that have a connection of type conclusion, reason, condition or consequence to the surrounding discourse units.…”

VisInReport: Complementing Visual Discourse Analytics Through Personalized Insight Reports

“…Altogether, we include 53 linguistic features, grouped into four categories, which can present the dynamics of deliberation (see [50] for reference). These features are extracted via the parsing pipeline developed by Hautli-Janisz et al [51]. Via rule-based methods we extract different types of questions [52].…”

Studying Visualization Guidelines According to Grounded Theory

“…Such a type of text data bares the characteristics of verbatim text, i.e., it is based on a social interactions, with a rapid creation cycle and evolution. In contrast to highly-edited text, commonly found in traditional print media, such as books and articles, this type of text brings along the challenge of being more noisy and heterogeneous [41]. Hence, to analyze such corpora, a tailored set of text analytics tools is required.…”