ALERT! The RMLMapper is rebuilt from the ground up on another repository (rmlmapper-java). All future development will now happen there. This repository remains available as legacy.
RML Data Retrieval considers different dataset and service descriptions and supports the RML Processor to access and retrieve the data sources based on those descriptions.
The access interfaces supported at the moment are
- Dump files
- DCAT for Data Catalogues (RML Data Retrieval considers Dataset and Distribution descriptions)
- Hydra for Web APIs (RML Data Retrieval considers IRI Template and Paged Collection descriptions)
- subset of D2RQ for Database Access Interface (RML Data Retrieval considers Database description)
- SPARQL-SD for SPARQL Service (RML Data Retrieval considers Dataset and Service descriptions)
- VoID for accessing RDF Datasets and/or their SPARQL endpoint (RML Data Retrieval considers Dataset and Dataset Description descriptions)
- CSVW for CSV files published on the Web (RML Data Retrieval considers Table and Dialect Description descriptions)
Details and examples of possible data source descriptions and their alignment with RML can be found at [RML Data Sources] (http://rml.io/RMLdataRetrieval.html).
Anastasia Dimou, Ruben Verborgh, Miel Vander Sande, Erik Mannens, Rik Van de Walle
Machine-Interpretable Dataset and Service Descriptions for Heterogeneous Data Access and Retrieval
In Proceedings of the 11th International Conference on Semantic Systems, SEMANTiCS2015
More information about the solution can be found at http://rml.io
This application is developed by Multimedia Lab http://www.mmlab.be
Copyright 2015, Multimedia Lab - Ghent University - iMinds
The RMLProcessor is released under the terms of the MIT license.