Management and Curation of EarthPortal Resources

Hello, and welcome to this project about the management and curation of the resources hosted in EarthPortal. The work was carried out within Data Terra (CNRS) during an internship from 1 September to 31 October 2026, supervised by Christelle Pierkot.

Each subfolder of this repository is dedicated to a specific activity and has its own README with more details:

• docs/: the reports and slides written during the internship, such as the quick reference table (cheatsheet) of the OntoPortal Metadata Guidelines, which the pipeline reads.
• earthportal_sa_audit/: the pipeline. Five notebooks build a state of the art of the semantic artefacts hosted in EarthPortal at the date of a snapshot (O'FAIRe scores, comparison with the Metadata Guidelines), and write one audit per artefact, as a page and as a JSON file. The notebooks also save the figures used in the slides (earthportal_sa_audit/figures/), so each figure comes from the committed data.

To learn more about a subfolder, please open its README first. Three short files at the root complete them:

• ARCHITECTURE.md: how the pipeline is built, and which file each step reads and writes;
• ADAPTING.md: what to change and check to run the pipeline on another OntoPortal portal;
• CHANGELOG.md: what changed in each release.

Reproducibility

1. Prerequisites

Our development environment runs smoothly on:

• Ubuntu 24.04.3 LTS (required: or any modern Linux system)
• Python 3.12.3 (required: 3.12 or later; check with python3 --version)
• Git 2.43.0 (required: check with git --version)
• Visual Studio Code 1.109.3 (optional: use any editor you like!)

You also need an API key of the portal you audit (EarthPortal by default): create an account on earthportal.eu, then copy the key shown on your account page. Only notebook 01 (data collection) uses it.

2. First-time setup

1. Clone the repository and move into it:

   git clone https://github.com/EarthPortal/SA_Management.git
   cd SA_Management

2. Create and activate a virtual environment:

   python3 -m venv .venv
   source .venv/bin/activate

3. Install the Python dependencies:

   pip install -r requirements.txt

4. Register the virtual environment as a Jupyter kernel:

   python3 -m ipykernel install --user --name data-terra-venv --display-name "Data Terra (.venv)"

Why? The notebooks expect a kernel named data-terra-venv. JupyterLab and jupyter nbconvert need this name (they are not in requirements.txt: install them with pip install jupyterlab). In VS Code, you can also choose .venv directly when you select the kernel (Python Environments > .venv).

5. Create your .env file from the template, then write your API key after PORTAL_APIKEY=:

   cp .env.example .env
   nano .env        # or open it in VS Code

Note: .env holds your personal API key. It is listed in .gitignore: never commit it and never share it. Besides the key, .env can hold RUN_DATE (the snapshot to work on) and, for another portal, the variables described in ADAPTING.md.

3. Updating an existing copy

If you already set up the project and only want the latest changes (for example, a new folder you did not have before):

1. Move to the project directory:

   cd SA_Management

2. Pull the latest changes:

   git pull

3. Activate your virtual environment:

   source .venv/bin/activate

4. Install any new dependencies:

   pip install -r requirements.txt

5. Compare .env.example with your .env, and add any new variable it lists.

Then open the README of the new folder to learn how to use it.

How to cite

If you use these notebooks or reports, please cite them with the DOI above. It always points to the latest release on Zenodo; each release also has its own DOI, listed on the Zenodo page. You can also click Cite this repository on the GitHub page.

The slides have their own record on Zenodo: . Each presentation is a version of this record, with its own DOI: to refer to one talk, please cite the DOI of its version.

License

MIT.