This repository contains code for the data integration of the BELTRANS project which studies Intra-Belgian translation flows between French and Dutch in the period 1970-2020. Data from different heterogenous data sources is integrated to create a FAIR corpus, this includes XML files from different sources but also existing large RDF dumps.
Preprocessing scripts for the different data sources are stored in the respective data-source folder. This mainly includes Python scripts but also RML mapping documents. The data integration is currently controlled by a bash script in the data-integration folder.
The integrated data is available in different formats on Zenodo 
During the project we have stored the RDF data in seperate named-graphs to indicate the provenance of the data,
e.g. data from KBR in the graph http://kbr-syracuse. The following queries can be used to retrieve RDF representations of the data
If + provenance is indicated, this means that the previously mentioned data source graphs, including RDF of bibliographic/authority data of KBR, BnF, KB and Unesco Index Translationum are exported as well.
- All integrated data flattened (all in a single named-graph): get-flattened-data
- All integrated data + provenance flattened (all in a single named-graph): get-flattened-data-with-provenance
- Due to limitations of the SPARQL 1.1 standard, it is not possible to CONSTRUCT a named-graph from a named-graph in the WHERE clause (w3c-cg/sparql-dev#31).
All integrated data: get-dataAll integrated data + provenance: get-data-with-provenance