Description
Metadata from datasets published on http://data.gouv.fr, converted to RDF, the [Semantic Web] model (https://en.wikipedia.org/wiki/Web_s%C3%A9mantic). This includes the metadata of the datasets hosted on data.gouv.fr, as well as those of the datasets that are only referenced on data.gouv.fr and hosted on other portals.
Given the exponential growth of data published on data.gouv.fr, the volume of metadata resulting from the transformation to RDF has exceeded what my server can process, especially in terms of disk space. This automatic conversion is therefore suspended.
Formats
Metadata is available in several formats:
- CSV (original format, see wiki)
- RDF Turtle (definition)
- RDF N-Quads (definition)
Data Model
The data model, just to find your way around. Properties not available via DCAT or Dublin Core have been added to a new ontology, DGFR (supports "content negotiation").
Formulation of queries
A SPARQL query editor and some examples are available [here] (http://www.data.maudry.com/query/en.html).
You can also make REST queries at http://www.data.maudry.com:3030/datagouvfr/sparql. In GET, the request must be [encoded] (http://meyerweb.com/eric/tools/dencoder/) and passed via the query parameter.
The metadata VoID is available in the main graph.
Update
I created a script that automates the tasks needed to convert and publish metadata to RDF. It is available on Github. This script is executed every two hours.
You can check the freshness of the data with [list of latest updated datasets] (https://www.data.maudry.com/en/datasets?_sort=-modified&_page=0&_pageSize=20).
Contact
If you have any questions:
Known issues
- The conversion script does not yet handle the tags of the datasets
Resources
| Name | Format | Description | Link |
|---|---|---|---|
| OpenRefine recipe for CSV > RDF conversion | TXT | Open resource | |
| Dataset Metadata and Resources (RDF) | Unknown | Open resource | |
| Ontology DGFR | Unknown | Open resource |