A5080674921- Natural Language Processing Techniques
- Topic Modeling
- Text Readability and Simplification
- Speech and dialogue systems
- Semantic Web and Ontologies
- Hate Speech and Cyberbullying Detection
- Multimodal Machine Learning Applications
- Wikis in Education and Collaboration
- Multi-Agent Systems and Negotiation
- Advanced Text Analysis Techniques
- Advanced Graph Neural Networks
- Speech Recognition and Synthesis
- Software Engineering Research
- Computational and Text Analysis Methods
- Culinary Culture and Tourism
- Interpreting and Communication in Healthcare
- Categorization, perception, and language
- Language, Metaphor, and Cognition
- Explainable Artificial Intelligence (XAI)
- Neural Networks and Reservoir Computing
- Religion, Society, and Development
- Biblical Studies and Interpretation
- Data Visualization and Analytics
- Language and cultural evolution
- Social Media and Politics
University of Copenhagen
2019-2024
IBM Research - Haifa
2014-2021
Uppsala University
2020-2021
University of Trento
2021
John Brown University
2021
Hebrew University of Jerusalem
2018-2020
Charles University
2019-2020
University of Oslo
2019-2020
University of Massachusetts Amherst
2020
Amherst College
2020
The recent release of ChatGPT has garnered widespread recognition for its exceptional ability to generate human-like conversations. Given usage by users from various nations and training on a vast multilingual corpus that includes diverse cultural societal norms, it is crucial evaluate effectiveness in adaptation. In this paper, we investigate the underlying background analyzing responses questions designed quantify human differences. Our findings suggest that, when prompted with American...
We describe a novel and unique argumentative structure dataset.This corpus consists of data extracted fro m hundreds Wikipedia articles using meticulously monitored manual annotation process.The result is 2,683 argument elements, collected in the context 33 controversial topics, organized under simp le claim-evidence structure.The obtained are publicly available for academic research.
We present the first parser for UCCA, a cross-linguistically applicable framework semantic representation, which builds on extensive typological work and supports rapid annotation. UCCA poses challenge existing parsing techniques, as it exhibits reentrancy (resulting in DAG structures), discontinuous structures non-terminal nodes corresponding to complex units. To our knowledge, conjunction of these formal properties is not supported by any parser. Our transition-based parser, uses novel...
Daniel Hershcovich, Stella Frank, Heather Lent, Miryam de Lhoneux, Mostafa Abdou, Stephanie Brandl, Emanuele Bugliarello, Laura Cabello Piqueras, Ilias Chalkidis, Ruixiang Cui, Constanza Fierro, Katerina Margatina, Phillip Rust, Anders Søgaard. Proceedings of the 60th Annual Meeting Association for Computational Linguistics (Volume 1: Long Papers). 2022.
The ability to consolidate information of different types is at the core intelligence, and has tremendous practical value in allowing learning for one task benefit from generalizations learned others. In this paper we tackle challenging improving semantic parsing performance, taking UCCA as a test case, AMR, SDP Universal Dependencies (UD) auxiliary tasks. We experiment on three languages, using uniform transition-based system architecture all Despite notable conceptual, formal domain...
Stephan Oepen, Omri Abend, Jan Hajic, Daniel Hershcovich, Marco Kuhlmann, Tim O'Gorman, Nianwen Xue, Jayeol Chun, Milan Straka, Zdenka Uresova. Proceedings of the Shared Task on Cross-Framework Meaning Representation Parsing at 2019 Conference Natural Language Learning. 2019.
Stephan Oepen, Omri Abend, Lasha Abzianidze, Johan Bos, Jan Hajic, Daniel Hershcovich, Bin Li, Tim O'Gorman, Nianwen Xue, Zeman. Proceedings of the CoNLL 2020 Shared Task: Cross-Framework Meaning Representation Parsing. 2020.
Pretrained language models have been shown to encode relational information, such as the relations between entities or concepts in knowledge-bases — (Paris, Capital, France). However, simple of this type can often be recovered heuristically and extent which implicitly reflect topological structure that is grounded world, perceptual structure, unknown. To explore question, we conduct a thorough case study on color. Namely, employ dataset monolexemic color terms chips represented CIELAB, space...
We present the SemEval 2019 shared task on Universal Conceptual Cognitive Annotation (UCCA) parsing in English, German and French, discuss participating systems results. UCCA is a cross-linguistically applicable framework for semantic representation, which builds extensive typological work supports rapid annotation. poses challenge existing techniques, as it exhibits reentrancy (resulting DAG structures), discontinuous structures non-terminal nodes corresponding to complex units. The has...
Abstract Building upon the considerable advances in Large Language Models (LLMs), we are now equipped to address more sophisticated tasks demanding a nuanced understanding of cross-cultural contexts. A key example is recipe adaptation, which goes beyond simple translation include grasp ingredients, culinary techniques, and dietary preferences specific given culture. We introduce new task involving cultural adaptation recipes between Chinese- English-speaking cuisines. To support this...
The climate impact of AI, and NLP research in particular, has become a serious issue given the enormous amount energy that is increasingly being used for training running computational models. Consequently, increasing focus placed on efficient NLP. However, this important initiative lacks simple guidelines would allow systematic reporting research. We argue deficiency one reasons why very few publications report key figures more thorough examination environmental impact, present quantitative...
The recent release of ChatGPT has garnered widespread recognition for its exceptional ability to generate human-like responses in dialogue. Given usage by users from various nations and training on a vast multilingual corpus that incorporates diverse cultural societal norms, it is crucial evaluate effectiveness adaptation. In this paper, we investigate the underlying background analyzing questions designed quantify human differences. Our findings suggest that, when prompted with American...
While the widespread use of Large Language Models (LLMs) brings convenience, it also raises concerns about credibility academic research and scholarly processes. To better understand these dynamics, we evaluate penetration LLMs across workflows from multiple perspectives dimensions, providing compelling evidence their growing influence. We propose a framework with two components: \texttt{ScholarLens}, curated dataset human- LLM-generated content writing peer review for multi-perspective...
The main goal of argumentation mining is to analyze argumentative structures within an argument-rich document, and reason about their composition.Recently, there also interest in the task simply detecting claims (sometimes called conclusion) general documents.In this work we ask how set detected can be augmented further, by adding it negation each claim.This presents two NLP problems: automatically negate a claim, when such negated claim plausibly used.We present first steps into solving...
Yonatan Bilu, Ariel Gera, Daniel Hershcovich, Benjamin Sznajder, Dan Lahav, Guy Moshkowich, Anael Malet, Assaf Gavron, Noam Slonim. Proceedings of the 57th Annual Meeting Association for Computational Linguistics. 2019.
Simone Tedeschi, Johan Bos, Thierry Declerck, Jan Hajič, Daniel Hershcovich, Eduard Hovy, Alexander Koller, Simon Krek, Steven Schockaert, Rico Sennrich, Ekaterina Shutova, Roberto Navigli. Proceedings of the 61st Annual Meeting Association for Computational Linguistics (Volume 1: Long Papers). 2023.
Neuroscientists evaluate deep neural networks for natural language processing as possible candidate models how is processed in the brain. These are often trained without explicit linguistic supervision, but have been shown to learn some structure absence of such supervision (Manning et al., 2020), potentially questioning relevance symbolic theories modeling cognitive processes (Warstadt and Bowman, 2020). We across two fMRI datasets whether align better with brain recordings, if their...
Abstract Creoles represent an under-explored and marginalized group of languages, with few available resources for NLP research. While the genealogical ties between a number highly resourced languages imply significant potential transfer learning, this is hampered due to lack annotated data. In work we present CreoleVal, collection benchmark datasets spanning 8 different tasks, covering up 28 Creole languages; it aggregate novel development reading comprehension relation classification,...
The non-indexed parts of the Internet (the Darknet) have become a haven for both legal and illegal anonymous activity. Given magnitude these networks, scalably monitoring their activity necessarily relies on automated tools, notably NLP tools. However, little is known about what characteristics texts communicated through Darknet have, how well do off-the-shelf tools this domain. This paper tackles gap performs an in-depth investigation text in Darknet, comparing it to clear net website with...
Daniel Hershcovich, Omri Abend, Ari Rappoport. Proceedings of the 2019 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019.