A5013599940- Natural Language Processing Techniques
- Topic Modeling
- Biomedical Text Mining and Ontologies
- Semantic Web and Ontologies
- linguistics and terminology studies
- Text Readability and Simplification
- Advanced Text Analysis Techniques
- Machine Learning and Algorithms
- Artificial Intelligence in Healthcare
- Linguistics and Discourse Analysis
- Translation Studies and Practices
- Machine Learning in Healthcare
- Speech and dialogue systems
- Sentiment Analysis and Opinion Mining
- Algorithms and Data Compression
- Speech Recognition and Synthesis
- Electronic Health Records Systems
- Video Analysis and Summarization
- Text and Document Classification Technologies
- Misinformation and Its Impacts
- Web Data Mining and Analysis
- Machine Learning in Bioinformatics
- Domain Adaptation and Few-Shot Learning
- Interpreting and Communication in Healthcare
- Artificial Intelligence in Law
Université d'Orléans
2020-2024
Ansys (United States)
2024
Laboratoire Ligérien de Linguistique
2019-2022
Centre National de la Recherche Scientifique
2018-2022
East Stroudsburg University
2022
Brandeis University
2022
RMIT University
2022
Dalle Molle Institute for Artificial Intelligence Research
2022
University of Zurich
2022
Mohamed bin Zayed University of Artificial Intelligence
2022
Anne-Lyse Minard, Manuela Speranza, Eneko Agirre, Itziar Aldabe, Marieke van Erp, Bernardo Magnini, German Rigau, Rubén Urizar. Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015). 2015.
In this article, we describe a system that reads news articles in four different languages and detects what happened, who is involved, where when. This event-centric information represented as episodic situational knowledge on individuals an interoperable RDF format allows for reasoning the implications of events. Our covers complete path from unstructured text to structured knowledge, which defined formal model links interpreted textual mentions things their representation instances. The...
This paper describes the approaches authors developed while participating in i2b2/VA 2010 challenge to automatically extract medical concepts and annotate assertions on relations between concepts.The authors'approaches rely both rule-based machine-learning methods. Natural language processing is used features from input texts; these are then authors' approaches. The Conditional Random Fields for concept extraction, Support Vector Machines assertion relation annotation. Depending task, tested...
In this paper, we discuss a cross-document coreference annotation schema that developed to further automatic extraction of timelines in the clinical domain.Lexical senses and choices are determined largely by context, but work requires reasoning across contexts not necessarily coherent.We found an approach relies less on context-guided annotator intuitions more schematic rules was most effective creating meaningful consistent relations.
In this paper we present a complete framework for the annotation of negation in Italian, which accounts both scope and focus, also language-specific phenomena such as negative concord. our view, complements more comprehensive Natural Language Processing tasks, temporal information processing sentiment analysis. We applied proposed guidelines built on top it to written texts, namely news articles tweets, thus producing annotated data total over 36,000 tokens.
This paper presents the work produced by students of University Orlans Masters in Natural Language Processing program way participating SemEval Task 6, LegalEval, which aims to enhance capabilities legal professionals through automated systems. Two out three sub-tasks available – Rhetorical Role prediction (RR) and Legal Named Entity Recognition (L-NER) were tackled, with express intent developing lightweight interpretable For L-NER sub-task, a CRF model was trained, augmented...
The HLT-FBK system is a suite of SVMsbased classification models for extracting time expressions, events and temporal relations, each with set features obtained the NewsReader NLP pipeline.HLT-FBK's best runs ranked 1st in all three domains, recall 0.30 over domains.Our attempts on increasing by considering SRL predicates as well utilizing event co-reference information links result significant improvements.