Advances in Information Extraction and Knowledge Graphs

This talk presents recent advances for extracting Knowledge Graphs from text using two core NLP tasks: Named Entity Recognition and Relation Extraction, with a special focus on Arabic. I will highlight state-of-the-art tools and datasets, and demonstrate how these components work together in practical extraction pipelines. In the second part, I introduce an Information Extraction Ontology designed to unify outputs from multiple systems and ensure semantic consistency, including schema.org and Wikidata. Finally, I show how this ontology can be embedded directly into AI prompts, enabling portable and more efficient Knowledge Graph construction within large language model workflows.