Description

Several important activities in the Oil and Gas (O&G) production domain can benefit from the use of Artificial Intelligence, especially large language models, for information retrieval from various databases. For data to be used coherently, it must be contextualized, thus enabling the association of information from Operation systems and corporate databases. The use of domain ontologies enables data contextualization based on the common definition of entities, their properties, and connections, facilitating information retrieval with greater reliability and precision.
OntoKG seeks the development of a domain ontology applied to an O&G production use case and the exploration of the domain ontology and knowledge graphs for information retrieval and contextualization to improve the fine-tuning and retrieval-augmented generation (RAG/GraphRAG) processes in LLM applications. Moreover, the project also seeks the development of machine learning-based methods for automating the instantiation of the developed domain ontology

Project Areas

Academic Partners

Industrial Partners