Management and completion of dynamic Knowledge Graph

ended

In MOKA, we will design and implement methods for managing the evolution of large and dynamic knowledge graphs (KG). Currently, KG are getting much attention from both industry and academia because they have properties that support the design and development of more intelligent and intelligible systems by connecting and structuring knowledge, ensuring semantic interoperability between information systems and providing richer content for automatic reasoning. However, one of the most widely agreed-upon problems experienced by data scientists and knowledge engineers is the maintenance of such KG over time. Since KG model the knowledge of a given domain, they have to be frequently updated to reflect the evolution of that domain which demand significant effort from experts and once updated changes are lost which leads to a general impoverishment of knowledge over time. To do so, our methods and tools will allow to (i) identify and characterize changes that occurs in KG when the domain evolves which will significantly reduce the efforts of domain experts in the KG maintenance tasks and (ii) define a mechanism to represent the evolution of a domain within a so-called Historical Knowledge Graph (HKG) that will allow users having a clear and deep understanding of the evolution of the domain knowledge over time. The proposed methods and tools will be evaluated on the KG of the Open Research Knowledge Graph initiative and the data of the space domain. We will have a significant impact on scientific, technological, economic and societal aspects. This will be made possible thanks to the proposed dissemination, exploitation and communication plans aimed at a large audience that will facilitate the development of the career of the researcher.