Composable Semantic Data Transformation Pipelines with Chimera

Marco Grassi, Mario Scrocca, Alessio Carenini, Marco Comerio, Irene Celino

Keywords: Semantic conversion, Declarative mappings, Data interoperability

TL;DR: Chimera minimises the effort required to specify custom semantic data transformation pipelines, offering different ready-to-use components to manipulate knowledge graphs and execute declarative mapping rules for their construction and exploitation.

Abstract: In a multi-stakeholder ecosystem, data exchange is not sufficient and semantic interoperability should be achieved to ensure different information systems can communicate without loss of meaning. A semantic conversion procedure relying on a reference ontology can enable an efficient conversion between data formats sharing common semantics, whilst generating an integrated and interoperable knowledge graph. The Chimera framework proposes a flexible and configurable solution to address different requirements for the integration of semantic converters across heterogeneous systems. Chimera minimises the effort required to specify custom semantic data transformation pipelines, offering different ready-to-use components to integrate heterogeneous data sources, manipulate knowledge graphs, and execute declarative mapping rules for their construction and exploitation. We present the Chimera framework as a resource for the Semantic Web community and we demonstrate its usage considering a challenging use case in the transportation domain.