Improving the Quality of Knowledge Graphs

Semantic technologies, including knowledge graphs, are technologies that are used in different computer science contexts and application areas (e.g., health, crime analysis, materials design) and e-science in general.

They are particularly important for search and integration of data sources, and are also one of the major technologies to make data FAIR (Findable, Accessible, Interoperable, Reusable).Semantically-enabled applications need high-quality knowledge graphs. Knowledge graphs with defects, although often useful, lead to problems when used in semantically-enabled applications.

Wrong conclusions may be derived or valid conclusions may be missed.

However, ensuring knowledge graphs quality, in particular regarding completeness (all relevant information is modeled) and correctness (no wrong information in the knowledge graphs), is a major challenge.

Current approaches for dealing with defects do not deal well with large knowledge graphs and the combination of correctness and completeness, and do not support domain experts well in repairing.In this project we will formalize, study and develop algorithms and tools for repairing knowledge graphs.

As a leading group in this field, we have excellent possibilites for addressing these challenges and provide important and necessary contributions to knowledge engineering.