ConCur: Knowledge Graph Construction and Curation

Knowledge Graphs (KGs), which are oft en understood to include both graph structured data composed of relational facts and ontological knowledge, have become increasingly popular for data management, knowledge representation and artifi cial intelligence. Such KGs oft en use RDF to represent facts and RDF Schema plus the Web Ontology Langauge (OWL) to represent ontologies. The project ConCur, hosted in University of Oxford, explores innovative theories and methods across various aspects for automating KG construction and curation, with a focus on ontology embedding, alignment, population and completion using both semantic technologies and machine learning methods, including (large) language models. The project has published a series of papers in Computer Science conferences and journals that report the developed methods, their experimental results and empirical studies; it has also constructed and released open-source data resources and system implementations that can benefit research and applications in both academia and industry. This report will summarize the project’s research results and briefl y introduce other project details including team members and industry partners.