Building an ontology of (dis) agreement space for argument mining

A major challenge for the automatic analysis of negotiation dialogues is that of modeling the (dis) agreement space shaped by different participants in the process of arguing. The evolution of argument practice is intimately tied with innovations in media: argument technologies augment the way collectives-groups, organizations, communities, societies-interact and reason in managing disagreement (Jackson, 2015). Current studies in argument mining have focused primarily on the automatic identification of the structure of arguments (eg, argument components and how these components are related to each other)(Somasundaran et al., 2007; Stab and Gurevych, 2014; Swanson et al., 2015; Feng and Hirst, 2011; Habernal and Gurevych, 2017; Peldszus and Stede, 2015). Bentahar et al.(2010) refer to these models of argumentation as monological or micro-level models, focusing on argument as product. In contrast, macro-level models (or dialogical models) and rhetorical models focus on the process of argument in a dialogue, being more suitable for analyzing negotiation or persuasion dialogues (Bentahar et al., 2010). The closest empirical work in argumentation mining community related to argumentation as process has been the detection of agreement and disagreement in online interactions (Abbott et al., 2011; Sridhar et al., 2015; Rosenthal and McKeown, 2015; Walker et al., 2012b). However, current methodologies for detection of agreement/disagreement do not point to what aspects the (dis) agreement has scope over; moreover, they do not allow us to predict if the related portions of texts are argumentative components (eg, claims and/or …