Résumé :
Knowledge graphs provide a structured and semantically meaningful way to represent and organize information using ontologies. In our research, we employ knowledge graphs to capture the relationships between entities within a specific application domain.
This presentation will introduce a preliminary project that demonstrates how knowledge graphs can enhance data journalism by incorporating basic reasoning into information extraction tasks such as entity linking.
Furthermore, we will explore how knowledge graph embedding can be utilized for automatic graph alignment without requiring prior knowledge.
We will delve into Ontology-based data management (OBDM) and demonstrate how it can be optimized for data querying, enabling efficient knowledge graph querying.
Finally, we will showcase a use case for shape representation and retrieval using ontologies.
Bio :
Cheikh Brahim EL VAIGH is an associate professor at the University of burgundy(France), CIAD Lab. He is also a visiting researcher at the Osaka University Institute for Datability Science since 2019. He received his MASc degree in computer science from the univ-rennes1 University (France) in 2017 and his Ph.D. in January 2021 from the same university. During his Ph.D., he worked on the joint use of knowledge graph and NLP for data-journalism at the INRIA/IRISA lab (Rennes, France). His research is focused on data analysis with join approaches leveraging knowledge graphs such as text and graph, or image and graph learning. He is also working on ontology mediated query answering querying with both RDF and description logics.