GerCCT: An Annotated Corpus for Mining Arguments in German Tweets on Climate Change
14464Authors
nan
Abstract
AbstractWhile the field of argument mining has grown notably in the last decade, research on the Twitter medium remains relatively understudied. Given the difficulty of mining arguments in tweets, recent work on creating annotated resources mainly utilized simplified annotation schemes that focus on single argument components, i.e., on claim or evidence. In this paper we strive to fill this research gap by presenting GerCCT, a new corpus of German tweets on climate change, which was annotated for a set of different argument components and properties. Additionally, we labelled sarcasm and toxic language to facilitate the development of tools for filtering out non-argumentative content. This, to the best of our knowledge, renders our corpus the first tweet resource annotated for argumentation, sarcasm and toxic language. We show that a comparatively complex annotation scheme can still yield promising inter-annotator agreement. We further present first good supervised classification results yielded by a fine-tuned BERT architecture.
Keywords
Full Content
Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 613–622 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 613 Integrating a Phrase Structure Corpus Grammar and a Lexical-Semantic Network: the HOLINET Knowledge Graph Jean-Philippe Prost Laboratoire Parole et Langage (LPL), Aix-Marseille Universit´e, France Jean-Philippe.Prost@univ-amu.fr Abstract In this paper we address the question of how to integrate grammar and lexical-semantic knowledge within a single and homogeneous knowledge graph. We introduce a graph modelling of grammar knowledge which enables its merging with a lexical-semantic network. Such an integrated representation is expected, for instance, to provide new material for language- related graph embeddings in order to model interactions between Syntax and Semantics. Our base model relies on a phrase structure grammar. The phrase structure is accounted for by both a Proof-Theoretical representation, through a Context-Free Grammar, and a Model-Theoretical one, through a constraint-based grammar. The constraint types colour the grammar layer with syntactic relationships such as Immediate Dominance, Linear Precedence, and more. We detail a creation process which infers the grammar layer from a corpus annotated in constituency and integrates it with a lexical-semantic network through a shared POS tagset. We implement the process, and experiment with the French Treebank and the JeuxDeMots lexical-semantic network. The outcome is the HOLINET knowledge graph. Keywords: Knowledge Graph, Lexical-Semantic Network, Grammar, Phrase Structure, Constituency, JeuxDeMots 1. Introduction While the pipeline software architecture, which steps from one linguistic dimension to the next, has been typ- ical for decades for most Natural Language Processing (NLP) applications, it often prevents many potential in- teractions across dimensions from actually occurring. A variety of sentence-level ambiguities, for instance, require the full sentence to be parsed morphologically, then syntactically, then semantically, prior to being dis- ambiguated through a pipeline. We hypothesize that a holistic modelling of the linguistic knowledge, which also approaches language as a whole rather than solely as a sum of its parts on various dimensions, would ease its processing and improve its performances in many respects. Knowledge graphs provide a convenient means for het- erogeneous knowledge to interact rather seamlessly. They open perspectives with regard to holistic ap- proaches to the modelling of knowledge, and linguistic knowledge in particular. Yet, the modelling of knowledge, in ways that make it suitable for Knowledge Graphs, is not always trivial and often constitutes a challenge as such. Concerning the representation of natural language syn- tax, the landscape of knowledge graphs has few re- sources, if any. Syntactic knowledge remains nearly exclusively represented through treebanks and other kinds of annotated corpora, and to a lesser ex- tent through knowledge-based grammars, such as the HPSG grammars from the DELPH-IN consortium1, or meta-grammars such as FRMG (de La Clergerie, 2005). Ultimately, our aim is to design and build a knowledge 1http://www.delph-in.net graph which overcomes representational discrepancies among knowledge areas that are relevant to natural lan- guage processing and make natural language process- able in a holistic manner. Assuming such a holistic knowledge graph, we may wonder …
Content truncated to 500 words
Labeled Text
AbstractWhile the field of argument mining has grown notably in the last decade, research on the Twitter medium remains relatively understudied. Given the difficulty of mining arguments in tweets, recent work on creating annotated resources mainly utilized simplified annotation schemes that focus on single argument components, i.e., on claim or evidence. In this paper we strive to fill this research gap by presenting GerCCT, a new corpus of German tweets on climate change, which was annotated for a set of different argument components and properties. Additionally, we labelled sarcasm and toxic language to facilitate the development of tools for filtering out non-argumentative content. This, to the best of our knowledge, renders our corpus the first tweet resource annotated for argumentation, sarcasm and toxic language. We show that a comparatively complex annotation scheme can still yield promising inter-annotator agreement. We further present first good supervised classification results yielded by a fine-tuned BERT architecture.
Publication Details
- Paper ID:
- 14464
- Year:
- 2022
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan