Papers published at EMNLP

We have two papers accepted at the EMNLP conference, as Findings papers:

Enrica Troiano, Sofie Labat, Marco Antonio Stranisci, Rossana Damiano, Viviana Patti, and Roman Klinger. Dealing with controversy: An emotion and coping strategy corpus based on role playing. In Findings of the Association for Computational Linguistics: EMNLP 2024. Association for Computational Linguistics, 2024. [ bib | arXiv | http | .pdf ]

Aswathy Velutharambath, Amelie Wuehrl, and Roman Klinger. How entangled is factuality and deception in german? In Findings of the Association for Computational Linguistics: EMNLP 2024. Association for Computational Linguistics, 2024. [ bib | arXiv | http | .pdf ]

The paper "Dealing with controversy: An emotion and coping strategy corpus based on role playing" starts out at the observation that emotion theories from psychology only received limited attention in natural language processing. The authors focus here on coping strategies, and ask workers on a crowdsourcing platform to act as they prefer to tackle challenging situations either by contacting, attacking, distancing, or rejecting the reason. The workers were then asked to described their emotion, appraisal, and coping approach from the role-played character or their own strategy. This lead to a corpus annotated with all these variables. 

The paper "How entangled is factuality and deception in german" deals with the question if deceptive texts that refer to factual information are different than deceptive texts that deal with non-factual information. The authors build on top of their previous work on a corpus annotated with factuality and deception variables, and model these variables based on the text. The detection of the deception variables turn out to be challenging, but interestingly, some fact-checking models are more successful with detecting the veracity of a statement based on honest texts than based on deceptive texts.