BibTeX

@inProceedings{lindahl-2024-disagreement-341074,
	title        = {Disagreement in Argumentation Annotation},
	abstract     = {Disagreement, perspective or error? There is a growing discussion against the idea of a unified ground truth in annotated data, as well as the usefulness of such a ground truth and resulting gold standard. In data perspectivism, this issue is exemplified with tasks such as hate speech or sentiment classification in which annotators’ different perspectives are important to include. In this paper we turn to argumentation, a related field which has had less focus from this point of view. Argumentation is difficult to annotate for several reasons, from the more practical parts of deciding where the argumentation begins and ends to questions of how argumentation is defined and what it consists of. Learning more about disagreement is therefore important in order to improve argument annotation and to better utilize argument annotated data. Because of this, we examine disagreement in two corpora annotated with argumentation both manually and computationally. We find that disagreement is often not because of annotation errors or mistakes but due to the possibility of multiple possible interpretations. More specifically, these interpretations can be over boundaries, label or existence of argumentation. These results emphasize the need for more thorough analysis of disagreement in data, outside of the more common inter-annotator agreement measures.},
	booktitle    = {3rd Workshop on Perspectivist Approaches to NLP, NLPerspectives 2024 at LREC-COLING 2024 - Workshop Proceedings},
	author       = {Lindahl, Anna},
	year         = {2024},
	ISBN         = {9782493814234},
}

@article{lindahl-borin-2024-annotation-333043,
	title        = {Annotation for computational argumentation analysis: Issues and perspectives},
	abstract     = {Argumentation has long been studied in a number of disciplines, including several branches of linguistics. In recent years, computational processing of argumentation has been added to the list, reflecting a general interest from the field of natural language processing (NLP) in building natural language understanding systems for increasingly intricate language phenomena. Computational argumentation analysis – referred to as argumentation mining in the NLP literature – requires large amounts of real-world text with manually analyzed argumentation. This process is known as annotation in the NLP literature and such annotated datasets are used both as “gold standards” for assessing the quality of NLP applications and as training data for the machine learning algorithms underlying most state of the art approaches to NLP. Argumentation annotation turns out to be complex, both because argumentation can be complex in itself and because it does not come across as a unitary phenomenon in the literature. In this survey we review how argumentation has been studied in other fields, how it has been annotated in NLP and what has been achieved so far. We conclude with describing some important current and future issues to be resolved.},
	journal      = {Language and Linguistics Compass},
	author       = {Lindahl, Anna and Borin, Lars},
	year         = {2024},
	volume       = {18},
	number       = {1},
}