Skip to content

Ontology Engineering with XD

For creating the ontology behind the Viewsari KG, our team follows the XD methodology, also reusing and proposing ODPs.

Requirements Engineering

Requirements engineering is the process of gathering, analyzing, and formalizing requirements for a system (an ontology / KG in our case).

User Interviews

From user interviews with five stakeholders from the art history domain, we gathered needs and wishes for Viewsari. We collected those in personas and adjacent user stories.

User Stories, Personas

We created four personas representing potential stakeholders. They have user scenarios and user requirements, as well as a set of competency questions. These parts are not rigid and can be adjusted with progress in the project. viewsari_personas (1189 x 841 mm).png

Mapping to ODPs

In order to increase the interoperability of our ontology, we matched the CQs to existing ODPs (and also created new ones). We reuse the Persons and Location CP from submissions in the ODP repository. Most importantly, we use the Provenance CP as inspiration for the representation of source evidence for extracted entities and cooccurrences. The solution in the repository does not fully match our use case, so we adapt it accordingly. We propose a cooccurrence class that has provenance information (e.g. a paragraph) and connects two or more resources. The examples below showcase what this would look like with data. We also propose a shortcut for authors and their connection to the cooccurrence. odp3.png odp2.png odp1.png

Formalization

In the formalization step, we construct a first foundational ontology for Viewsari. This ontology includes the following classes: odp1.png For a better view, please open the file in the directory

An exemplary view of how cooccurrences are connected to the work level: odp1.png

References

[1] E. Blomqvist, K. Hammar, V. Presutti, Engineering Ontologies with Patterns - The eXtreme Design Methodology., in: Ontology Engineering with Ontology Design Patterns - Founda- tions and Applications, 2016, pp. 23–50. doi:10.3233/978-1-61499-676