Creating Virtual Knowledge Graphs from Software-Internal Data
Maximilian Weigand, Alexander Fay · IECON 2022 – 48th Annual Conference of the IEEE Industrial Electronics Society · 2022
Due to their capability to model information about various domains in a semantically rich way, knowledge graphs gain importance in the engineering of automated systems. Still, the integration of engineering artifacts, which are usually created by software, is a challenge, due to missing interfaces of engineering software. In this paper, we propose a method for extracting information directly from running engineering software. The core concept of this method are virtual graphs, which do not physically store data, but instead have capabilities to access data from the software’s internal information model on demand. This information can then be accessed by other engineering software, which avoids typical, tedious workflows consisting of data exports and imports from and to different tools. For this concept, a generic concept and an architecture for a landscape of engineering tools are presented. They are validated by a prototypical implementation, which creates a virtual graph of the internal data of a SysML-tool and makes it accessible by a graph database.