Inducing Commonsense Knowledge Using Vector Space Embeddings
- Author:
- Zied Bouraoui
- HDR Defended on :
- Nov 17, 2022 • salle des thèses, Faculté des sciences Jean Perrin
Summary
My habilitation provides a high-level overview of my contributions on inducing commonsense knowledge using vector space representations, with a focus on :
- Learning conceptual space representations (learning entity embeddings and region-based representations of concepts, learning interpretable dimensions)
- Modelling relational knowledge (relation induction in word embedding and pre-trained language models, learning of distributional relation vectors)
- Deriving high quality vectors from contextualised LMs and applications to few-shot learning.
- Plausible reasoning about ontologies (automated rule base completion, inconsistency handling and belief merging)
Committee
- Isabelle Bloch (Sorbonne Université)
- Jesse Davis (KU Leuven)
- Philippe Langlais (Université de Montréal)
- Daniel Le Berre (Université d’Artois)
- Henri Prade (IRIT Toulouse)
- Marie-Christine Rousset (Université Grenoble Alpes)
- Steven Schockaert (Cardiff University)