Intelligent Information Management

Knowledge graphs are used across various domains to represent information in a structured data model. In real-world applications, knowledge graphs evolve continuously as entities and relationships may be added, changed, or removed over time. Current approaches to visualize such changes primarily focus on presenting changes at the global graph-level and typically present these in a graph-based or tabular format. However, as graph size or update frequency increases, the resulting visualizations become difficult to manage. Furthermore, users often find the presentation of changes in graph or tabular format to be unintuitive and hard to interpret.

The objective of this thesis is to develop a user-friendly approach for visualizing the evolution of knowledge graphs at the entity-level. First, a requirements analysis needs to be conducted, and existing approaches to visualize knowledge graph evolution need to be reviewed. Existing solutions have to be classified into groups and assessed based on the identified requirements. Based on these findings, a concept for an entity-scoped approach to visualize knowledge graph evolution has to be developed. The feasibility of this approach has to be demonstrated through a proof-of-concept implementation and evaluated using a suitable methodology.