Project: Schema-centered knowledge graph visualization and visual querying
Description
Since Google introduced knowledge graphs (KG) in 2012, the interest in the topic has increased continuously, and many companies applied this or comparable techniques. Nowadays, KGs are used in various fields, from product recommendations for online shops to finding connections between criminals in criminal investigations. While automated methods for creating and correcting KGs support developers in structuring the data and building the graph, visualization is needed to get insights into the data, especially when the task is explorative , and it is not initially known what to look for. Furthermore, KG visualization is usually combined with queries to limit the size of the visualized network.
Nevertheless, state-of-the-art KG visualization approaches focus either on standard network visualization techniques (e.g., [1]), neglecting the underlying schema of KGs, or visualize the KG and the schema side by side (e.g., [2]).
In both approaches, queries are commonly complicated and can only be done by KG experts, while in general, a broader audience is interested in KG-related results.
Up to this point, visual queries are mainly investigated in the direction of visually constructing a query (e.g., as a graph [3]) but do not support a visual abstraction from the query itself.
Therefore, users are forced to formulate queries instead of visually defining what they are interested in.
In this project, we want to overcome this by exploring schema-centered approaches for visualizing KGs, leading to potential more scalable KG visualizations or more insight into the internals of KGs and their structure. Furthermore, the native embedding of the schema within the visualization should enable the formulation of visual queries, which non-KG experts can perform.
Depending on the found visualization, this project can also lead in the direction of
1) Visual comparison of query results
2) Visualizing changes in the KG for changes in the underlying schema
3) Visualizing missing data
4)Visualizing KGs for non-technical experts
Plenty of open KG datasets exist on the internet. One example that can be used as a starting point is https://deepgraphlearning.github.io/project/wikidata5m.
Requirements:
1) Good programming skills
2) Visualization background
3) Understanding of knowledge graphs
References:
[1] Nararatwong, Rungsiman, Natthawut Kertkeidkachorn, and Ryutaro Ichise. "Knowledge graph visualization: Challenges, framework, and implementation." 2020 IEEE Third International Conference on Artificial Intelligence and Knowledge Engineering (AIKE). IEEE, 2020.
[2] Borland D, Christopherson L, Schmitt C. Ontology-Based Interactive Visualization of Patient-Generated Research Questions. Appl Clin Inform. 2019 May;10(3):377-386. doi: 10.1055/s-0039-1688938. Epub 2019 Jun 5. PMID: 31167249; PMCID: PMC6551210.
[3] Pienta, Robert, et al. "Visage: Interactive visual graph querying." Proceedings of the International Working Conference on Advanced Visual Interfaces. 2016.
Details
- Supervisor
-
Kay Roggenbuck
- Secondary supervisor
-
Stef van den Elzen
- Interested?
- Get in contact