Semantic Knowledge Graphs for Distributed Data Spaces: The Public Procurement Pilot Experience

This paper presents the experience gained in the context of a European pilot project funded by the ISA2 programme. It aims at constructing a semantic knowledge graph that establishes a distributed data space for public procurement. We describe the results obtained, the follow up actions and the main lessons learnt from the construction of the knowledge graph. This latter requires to support different data gover- nance scenarios: some partners control, with their own tools, the building process of their portion of the knowledge graph. Other partners partici- pate in the pilot by providing only their open CSV/XML/JSON datasets, in which case transformations are required. These are performed on the infrastructure made available by the European Big Data Test Infrastruc- ture (BDTI). The paper introduces the design and implementation of the knowledge graph construction process within such a BDTI infras- tructure. By instantiating an OWL ontology created for this purpose, we are able to provide a declarative description of the whole workflow required to transform input data into RDF output data, which form the knowledge graph. The declarative description is therefore used to provide instructions to a workflow engine we use (Apache Airflow) for knowledge graph construction purposes.