PSMV-RDF

Note

This repository is under active development. Features, documentation, and structure will change frequently.

Plant Protection Products (PSMV) as Linked Data

A Python module for converting Swiss plant protection product data from CSV format to RDF and publishing it to the LINDAS Linked Data Service.

Reproduce the data integration pipeline

1. Set up the virtual environment

   If uv is not yet installed: curl -LsSf https://astral.sh/uv/install.sh | sh

   uv venv psmv-rdf
   source psmv-rdf/bin/activate  
   

2. Install the package in editable mode

   uv pip install -e .
   

3. Start the data integration pipeline

   python -m service.pipeline
   

4. To upload the graph, first, place a .env file in the directory root:

   LINDAS_USER=********
   LINDAS_PASSWORD=************
   ENDPOINT=https://graphdb.lindas.admin.ch/repositories/lindas/rdf-graphs/service
   GRAPH=https://lindas.admin.ch/fsvo/plant-protection-products
   

   Then trigger the upload to LINDAS:

   python -m service.upload_graph
   

Project Structure

psmv-rdf/
├── .github/
│   └── workflows/  # github actions
├── data/           # any non-RDF data files, used as input data
│   ├── mapping/    # yaml mapping files
│   ├── namespaces/ # 
│   ├── processed/  # 
│   └── mapping/    # input CSV files
├── docs/           # project documentation
├── rdf/
│   ├── ontology/   # OWL ontology documentation
│   ├── shapes/     # SHACL shapes, also used as data model documentation
│   ├── data/       # the actual RDF data, split by classes
│   ├── example/    # example turtle files used for reference
│   └── processed/  # any automatically written/derived/merged turtle files
├── services/
│   └── pipeline.py       
├── src/
│    ├── sparql     # SPARQL queries and inference rules
│    └── python/    # Python scripts for specific tasks
├── tests/
├── LICENSE
├── README.md
└── .gitignore

Documentation

All ontology documentation files are written to rdf/ontology. You may inspect a visual representation of the ontology used here.

A more restricted data model is written in SHACL and can be inspected here.

Note

We should align the SHACL data model with this documentation: https://github.com/user-attachments/files/28257903/177035312.Datenmodell.PSMV.docx

Project dependencies are listed in pyproject.toml.

Example queries

List some products

List all prodcuts containing sulphur

Which products can be used for sweet corn when thrips are the target pest?

Acknowledgments

• Built with rdflib
• Integrates with LINDAS, the Swiss federal linked data service.
• Orignial ontology and pipeline by Damian Oswald with plant protection pipeline