Argumentation Mining
The area of argument mining has traditionally been very focused on the domain of law, because arguments play a central role in reaching the conclusion of a case, but also for the practical reason that legal texts are relatively structured and the arguments quite explicit. Arguments in other types of reasoning in free text, particularly the precise relation between argumentative statements and between whole arguments, is typically quite implicit. Neither is it the case that arguments "in the wild" neatly proceed from premises to conclusion in chronological order.
General notes
What is argumentation? And what is an argument?
- Toulmin' argumentation model
- Besnard and Hunter, 2008: simple model: premises + inference -> conclusion
- Argument mining pipeline
Potential data for argument mining:
Papers
- Wyner (2010) Approaches to text mining arguments from legal cases
- Mochales et al. 2011 - Argumentation mining
- Sketches some nice use cases for the use of argument mining in case-based reasoning and information retrieval
- Lippi 2016 - Argumentation mining state of the art emerging trends
- Moens 2018 - Argumentation mining how can a machine acquire common sense and world knowledge
- Lippi 2015 - Context-independent Claim Detection for Argument Mining
- Claim mining
- Sardianos 2015-Argument extraction from news
- Used word embeddings to generate more domain-specific word features that are in the end used in a typical ML setup (slightly disappointing)
- Claim to be semi-supervised, but are actually not
- Ashley 2013 - Toward constructing evidence-based legal arguments using legal decision documents and machine learning
- Shulayeva 2017 - Recognizing cited facts and principles in legal judgements
- Used automatic labelling of facts and principles based on linguistic features
- Falakmasir 2017: Utilizing Vector Space Models for Identifying Legal Factors from Text
- Legal factors: "stereotypical patterns of fact that tend to strengthen or weaken a side's argument in a legal claim"
- Also see Westermann 2019
- Al-Abdulkarim 2016 contains a critique on the fact that factors are typically represented in a Boolean fashion, whereas they are not defined in Boolean terms.
- Knott 1994 -Using linguistic phenomena to motivate a set of coherence relations
- Contains list of features that are indicative of arguments
- Cited by multiple sources, e.g. Lippi 2015
- Heb ik vertaald!
- TODO: Ran Levy et al. Context dependent claim detection. In Jan Hajic and Junichi Tsujii, editors, COLING 2014, Dublin, Ireland, pages 1489–1500. ACL, 2014.
- Work behind IBM Debater; hopefully useful info on features for detecting premises, claims
Other TODOs for later:
Also, in my current approach I'm not really doing sentence classification but Argumentative Zoning. For that, see maybe:
- Simone Teufel. 1999. Argumentative zoning: Information extraction from scientific text. Technical report.
- Yufan Guo, Anna Korhonen, and Thierry Poibeau. 2011. A weakly-supervised approach to argumentative zoning of scientific documents. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP ’11, pages 273–283, Stroudsburg, PA, USA. Association for Computational Linguistics.
Other TODOs:
- [C] Formalized argument approach: Prakken, H., and Sartor, G. 2015. Law and Logic: A Review from an Argumentation Perspective. Artificial Intelligence 227: 214–245. doi.org/10.1016/j.artint.2015.06.005
- Almost all works I find on argumentation mining use the very easy working definition of argument by Walton 2009; see Douglas Walton. Argumentation theory: A very short introduction. In Guillermo Simari and Iyad Rahwan, editors, Argumentation in Artificial Intelligence, pages 1–22. Springer US, 2009.
- Moens 2007: Moens MF, Boiy E, Palau RM, Reed C (2007) Automatic detection of arguments in legal texts. In: Proceedings of the 11th international conference on Artificial intelligence and law. ACM, pp 225–230
- Probably nothing new; quite old
- Raquel Mochales Palau and Marie-Francine Moens. 2009. Argumentation mining: the detection, classification and structure of arguments in text. In Proceedings of the Twelfth International Conference on Artificial Intelligence and Law (ICAIL 2009),, pages 98–109. ACM.