Exploiting Debate Portals for Semi-Supervised Argumentation Mining in User-Generated Web Discourse

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 2127–2137, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. Exploiting Debate Portals for Semi-Supervised Argumentation Mining in User-Generated Web Discourse Ivan Habernal† and Iryna Gurevych†‡ †Ubiquitous Knowledge Processing Lab (UKP-TUDA) Department of Computer Science, Technische Universit¨at Darmstadt ‡Ubiquitous Knowledge Processing Lab (UKP-DIPF) German Institute for Educational Research www.ukp.tu-darmstadt.de Abstract Analyzing arguments in user-generated Web discourse has recently gained atten- tion in argumentation mining, an evolving field of NLP. Current approaches, which employ fully-supervised machine learn- ing, are usually domain dependent and suffer from the lack of large and diverse annotated corpora. However, annotating arguments in discourse is costly, error- prone, and highly context-dependent. We asked whether leveraging unlabeled data in a semi-supervised manner can boost the performance of argument component identification and to which extent is the approach independent of domain and reg- ister. We propose novel features that ex- ploit clustering of unlabeled data from de- bate portals based on a word embeddings representation. Using these features, we significantly outperform several baselines in the cross-validation, cross-domain, and cross-register evaluation scenarios. 1 Introduction Argumentation mining, an evolving sub-field of NLP, deals with analyzing argumentation1 in var- ious genres, such as legal cases (Mochales and Moens, 2011), student essays (Stab and Gurevych, 2014a), and medical and scientific articles (Green, 2014; Teufel and Moens, 2002). Recently, the fo- cus of argumentation mining has also shifted to the Web registers (such as comments to articles, forum posts, or blogs) which is motivated by the need of 1Argumentation is a verbal activity for which the goal consists of convincing the listener or reader of the acceptabil- ity of a standpoint by means of a constellation of propositions justifying or refuting the proposition expressed in the stand- point (van Eemeren et al., 2002) or the art of persuading others to think or act in a definite way, including all writing and speaking which is persuasive in form (Ketcham, 1917). retrieving and understanding ordinary people’s ar- guments to various contentious topics on the large scale. Applications include passenger rights and protection (Park and Cardie, 2014), hotel reviews (Wachsmuth et al., 2014), and controversies in ed- ucation (Habernal et al., 2014). Despite the plethora of existing argumentation theories (van Eemeren et al., 2014), the preva- lent view in argumentation mining treats argu- ments as discourse structures consisting of sev- eral argument components, such as claims and premises (Peldszus and Stede, 2013). Current approaches to automatic analysis of argumenta- tion usually follow the fully supervised machine- learning paradigm (Biran and Rambow, 2011; Stab and Gurevych, 2014b; Park and Cardie, 2014) and rely on manually annotated datasets. Only few publicly available argumentation cor- pora exist, as annotations are costly, error-prone, and require skilled human annotators (Stab and Gurevych, 2014a; Habernal et al., 2014). To overcome the limited scope and size of the existing annotated corpora, semi-supervised meth- ods can be adopted, as they gain performance by exploiting large unlabeled datasets (Settles, 2012). However, unlike …

Content truncated to 500 words