Overview of DialAM-2024: Argument Mining in Natural Language Dialogues

Argumentation is the process by which humans rationally elaborate their thoughts and opinions in written (e.g., essays) or spoken (e.g., debates) contexts. Argument Mining research, however, has been focused on either written argumentation or spoken argumentation but without considering any additional information, e.g., speech acts and intentions. In this paper, we present an overview of DialAM-2024, the first shared task in dialogical argument mining, where argumentative relations and speech illocutions are modelled together in a unified framework. The task was divided into two different sub-tasks: the identification of propositional relations and the identification of illocutionary relations. Six different teams explored different methodologies to leverage both sources of information to reconstruct argument maps containing the locutions uttered in the speeches and the argumentative propositions implicit in them.

Argument Mining (Lawrence and Reed, 2020) investigates the automatic extraction of argument structures from natural language inputs. The nature of argumentation, however, can be very variable depending on its context, presenting significant differences between written and spoken argumentation (Hitchcock, 2009), and between monological and dialogical argumentation (O’Keefe, 1977). Research in argument mining has mainly focused on the extraction of arguments only considering argument annotations such as premises and claims (Stab et al., 2018; Reimers et al., 2019) or attacks and supports between propositions (Hou and Jochim, 2017; Ruiz-Dolz et al., 2021; Saadat-Yazdi et al., 2023),

Inference Anchoring Theory (IAT) was proposed as an annotation framework for dialogue argumentation where not only the structure of arguments is captured, but also the speech acts and speaker intent is also annotated to support and contextualise argumentation in dialogues

This way, the two DialAM-2024 sub-tasks are defined as follows:

A. Identification of Propositional Relations. In the first task, the goal is to detect argumentative relations existing between the argumentative propositions directly extracted from the locutions uttered in the argumentative dialogues. Such relations are: Inference, Conflict, and Rephrase.

B. Identification of Illocutionary Relations. In the second task, the goal is to detect illocutonary relations existing between locutions uttered in the dialogue and the argumentative propositions associated with them including: Asserting, Agreeing, Arguing, Disagreeing, Challenging, Restating, Pure Questioning, Rhetorical Questioning, and Assertive Questioning. The final goal of the DialAM-2024 shared task is, therefore, to reconstruct graph-structured argument maps, containing locutions and argument propositions previously identified and segmented from argumentative dialogues.