The resource

SDM-RDFizer introduces a novel set of operators to execute mapping rules in a data integration system; they allow for an efficient creation of knowledge graphs from heterogeneous data sources. Although the current version of SDM-RDFizer is customized for RML, the set of operators can be easily extended for other mapping rule languages and data models to represent knowledge graphs. Results of the experimental studies comparing the performance of SDM-RDFizer illustrate the novelty of the proposed work with the state of the art. We hope that these results encourage the community to advance existing approaches to scale up to the avalanche of available data that is expected in the next years.

Availability:

SDM-RDFizer is released publicly by the Scientific Data Management (SDM) group at TIB, Hannover. TIB is one of the largest libraries for science and technology in the world TIB, and following its policy of engaging open access to scientific artifacts, will keep available SDM-RDFizer as a tool for supporting the creation of knowledge graphs. The SDM-RDFizer is open source, written in Python 3, and available under the Apache License 2.0 license; it is regularly updated with new features. Additionally, following open science good practices, we register the tool at the Zenodo platform, which takes the Github repository and gives a general DOI to the engine and also a DOI for each release of the code. Thus, users and practitioners can use and cite a specific version of the engine, ensuring reproducibility and traceability of any experimental evaluation.

Utility:

A docker image of SDM-RDFizer is available at DockerHub and the Github repository of the resource, provides a detailed explanation of how to create and run the Docker container. Furthermore, the activity of commits in the Github repository evidence the attention paid to the creation of new versions, as well as to the addressing of the issues identified by the users of the tool.

Predicted Impact:

The number of visits of knowledge graphs like DBpedia and Wikidata, and the current developments in scientific and industrial areas evidence of the need of providing efficient tools knowledge graph management at scale. Results of experimental evaluations of the SDM-RDFizer illustrate the benefits of grounding solutions for the problem of knowledge graph creation in the well-established areas of data integration systems and query processing. Thus, we ambition that they will be the starting point of future developments, e.g., for the optimization and distribution of mapping rule executions, as well as for semantically enriching data integration systems whose execution enable the explainability of the whole process of knowledge graph creation.