معرفی
Matthieu Labeau is a Senior Lecturer at Télécom Paris, affiliated with the Department of Image, Data, Signal (IDS). He joined the institution in 2019 after completing his PhD at the University of Paris-Saclay and a postdoctoral position at the University of Edinburgh. His research primarily centers on Natural Language Processing (NLP), with specialized interests in representation learning, language modeling, and conversational AI.
His work spans:
- Core NLP: Contextual word representations, semantic alignment, and polysemy analysis.
- Machine Learning: Hierarchical classification, graph prediction, and few-shot learning techniques.
- Applications: Emotion recognition in dialogues, persuasiveness decoding, and educational NLP tools.
Labeau leads research in the Signal, Statistics and Learning (S2A) team at the Information Processing and Communication Laboratory (LTCI). His recent publications demonstrate a strong focus on improving language model interpretability and efficiency, with innovations in tokenization effects and multimodal fusion. Though no awards or grants are mentioned, his consistent output in top-tier venues (e.g., NeurIPS, ACL, AAAI) highlights significant scholarly contributions.
He actively collaborates on tools like EZCAT for conversation annotation and mentors researchers in NLP projects. Current work explores LLM capabilities in persuasion assessment and optimal transport methods for graph-based learning.
