Ontology Driven Transformation of Camunda BPMN Instances into RDF Knowledge Graphs

! AIKG-SD 2025 Summer School co-located with the NFDI4DS Conference 2025 November 25-26, 2025, Berlin, Germany 1! Ontology Driven Transformation of Camunda BPMN Instances into RDF Knowledge Graphs Florian Hahn and Sara Todorovikj Chemnitz University of Technology Abstract The increasing number of Open Government Data initiatives necessitates the implementation of transparent, interoperable, and automatable publication workflows. Business Process Model and Notation (BPMN) provides a standardized notation for modeling such workflows, while Resource Description Framework (RDF) representations enhance semantic interoperability and integration with Linked Data ecosystems. This research is situated within a broader project on best practices for Open Government Dataset publishing that are supported by BPMN, wherein each dataset publication is associated with a BPMN instance executed in Camunda. This integration enables the documentation and automation of metadata, provenance, and validation steps, thereby fostering trust and reproducibility within the publication pipeline. The central research question is: Can an ontology driven transformation of BPMN 2.0 processes executed in Camunda via REST API into RDF using the BPMN Based Ontology achieve semantic completeness and comparable query performance relative to Camunda’s native API? We hypothesize that a workflow which extracts the BPMN 2.0 model structure through Camunda REST API and converts it into RDF classes and relationships according to BBO produces a semantically complete, lossless representation. Under this hypothesis, in future, SPARQL queries on the resulting graph will reproduce at least 80 % of native query functionality and introduce no more than 20 % performance overhead. The methodology comprises: 1. Model Extraction: Use Camunda REST API to retrieve BPMN 2.0 definitions and runtime instance data. 2. Ontology Driven Transformation: Apply the BPMN Based Ontology (BBO) to map BPMN elements (tasks, gateways, events, pools, lanes, data objects) into RDF/OWL classes and properties. Implement transformations via XSLT, supplemented by RDFLib for graph construction and fine tuning. 3. Semantic Completeness Evaluation: Compare entities and relations encoded in the RDF graph against the native Camunda data model. Employ metrics such as recall and precision aiming for ≥95 % alignment, ensuring that SPARQL queries yield equivalent informational coverage to native API queries. 4. Performance Analysis: Design representative SPARQL queries and benchmark them against equivalent native Camunda queries in terms of execution time and resource consumption, targeting at most 20 % overhead. 5. Case Study: Open Government Dataset Publication: Model a concrete BPMN process for dataset publication (including metadata validation, license verification, versioning, and portal deployment), implement it in Camunda, and concurrently generate an RDF graph capturing process execution, metadata, and provenance according to PROV-O. ! AIKG-SD 2025 Summer School co-located with the NFDI4DS Conference 2025 November 25-26, 2025, Berlin, Germany 2! The expected outcomes include demonstrating that BPMN driven workflows, which will be available on GitHub later on, can be transformed losslessly into RDF knowledge graphs. This will enable them to be integrated into Linked Data infrastructures for enhanced analysis. Future work envisions integration into large scale Open Government Data platforms, extension to track BPMN model versioning and historical queries in RDF, application of semantic validation via SHACL on both data and process graphs, automated process optimization through pattern detection in the knowledge graph, and generalization to other domains. This poster presents a prototype implementation and initial evaluation results as a foundational component of a doctoral research agenda connecting process management and Open Data publication practices. References 1. A. Annane, N. Aussenac-Gilles, and M. Kamel, “BBO: BPMN 2.0 Based Ontology for Business Process Representation”, in 20th European Conference on Knowledge Management (ECKM 2019), Lisbon, Portugal, September 5–6, 2019. [Online]. Available: https://hal.science/hal-02365012v1/document [Accessed: June 10, 2025]. 2. Judie Attard, Fabrizio Orlandi, Simon Scerri, and Sören Auer, “A systematic review of open government data initiatives,” Government Information Quarterly, vol. 32, no. 4, pp. 399–418, August 2015. [Online]. Available: https://doi.org/10.1016/j.giq.2015.07.006 [Accessed: June 10, 2025]. 3. Camunda Services GmbH, “REST API Reference,” Camunda Documentation. [Online]. Available: https://docs.camunda.org/manual/latest/reference/rest/ [Accessed: June 10, 2025]. 4. Michael Schmidt, Michael Meier, and Georg Lausen, “Foundations of SPARQL Query Optimization,” in 13th International Conference on Database Theory (ICDT) , March 22, 2010. [Online]. Available: https://arxiv.org/abs/0812.3788 [Accessed: June 10, 2025]. 5. RDFLib contributors, “RDFLib Documentation,” Read the Docs. [Online]. Available: https://rdflib.readthedocs.io/ [Accessed: June 10, 2025]. 6. W3C, “PROV-O:ThePROVOntology,” W3CRecommendation. [Online]. Available: https://www.w3.org/TR/prov-o/ [Accessed: June 10, 2025]. 7. W3C, “Shapes Constraint Language (SHACL),” W3C Recommendation, October 8, 2015. [Online]. Available: https://www.w3.org/TR/shacl/ [Accessed: June 10, 2025]. 8. MDNWebDocs, “XSLT–ExtensibleStylesheetLanguageTransformations,”[Online]. Available: https://developer.mozilla.org/de/docs/Web/XML/XSLT [Accessed: June10,2025]. 9. RDFlib Developers, “RDFlib Documentation,” [Online]. Available: https://rdflib.readthedocs.io/en/stable/ [Accessed: June 10, 2025]. 10. Bundesverwaltungsamt, “Attribute Catalog Prozessmanagement Toolkit (PMT),” [Online].Available: https://www.bva.bund.de/SharedDocs/Downloads/DE/Behoerden/Beratung/Prozessma nagement/PMT/Attributkatalog_PMT.pdf?__blob=publicationFile&v=2 .[Accessed: June 10, 2025]. ! AIKG-SD 2025 Summer School co-located with the NFDI4DS Conference 2025 November 25-26, 2025, Berlin, Germany 3! 11. Camunda, “Working with APIs and Tools,” [Online]. Available: https://docs.camunda.io/docs/apis-tools/working-with-apis-tools/ [Accessed: June 10, 2025]. 12. Damaris Naomi Dolha, Robert Andrei Buchmann, “Experiments with natural language queries on RDF vs. XML-serialized BPMN diagrams,” Procedia Computer Science, vol. 202, pp. 3246–3255, 2024. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S187705092402338X [Accessed: June 10, 2025].