- M@TE models begin their life as github repositories, based on a this repository template
- New models are spawned using the github issues functionality, within this repository
- We provide workflows (using github actions) that aim to harvest and reuse as much existing metadata as possible, using persistent identifiers such as ORCiDs and RoRs
- The result is a model repository that comes with a rich metadata document, based on the RO-Crate project
- This process also assembles material so we can feature the model on the M@TE website
1) Make a new model request:
- open a new issue using the New model request
- fill in as much information as you can and submit the issue
- a
new model
New model label will appear on the original issue - This will trigger workflows that collect this information and source additional metadata relavant to your model
- You will see a summary report for your model, including potential missing information or errors (e.g. URLs that didn't resolve)
- If you have requested the content of your model to be embargoed, you should see an embargo requested label. We can still create a model repository, metadata file, and DOI for this model
2) Editing and review:
- By editing the first comment in the issue discussion, you will trigger a rebuild of the model information
- When you are satisfied with the information contained in the report, add a
review requested
Review requested label to the issue
3) Model repository creation:
- The model_reviewers team will add an
model approved
Model approved label to the issue - this triggers the creation of a new model repository within the M@TE organization, based on this simple template
- the model metadata is stored as an RO-Crate (named
ro-crate-metadata.json) in the top level of the repository - A
model created
Model created label will appear on your issue - The github account holder who submitted the model will have owner privillages.
4) Clone and configure your model repository:
- Your model repository is now yours to customize and fill with files (payload)
- We encourage you to use github to add any material that is within github's typical repository limits (files < 100 Mb, total repository size < 5 Gb)
- You may customize the model repository as you feel fit. This may require editing of the metadata file () to reflect changes in directory structure
- Push any changes back to the M@TE organization.
5) Upload model to our NCI Server:
- Add an upload to NCI label to the model submission issue
- Let us know (via the original issue) if you need to add additional files that exceed Github's limits
- The model_reviewers team will contact you via email to orchestrate this process
6) Model published:
- Your model will be published on the NCI GeoNetworkCatalog
- It will recieve a DOI and this will be automatically added to your model metadata
- Your model will be featured on the M@TE website
- a
model published
Model published label will appear on the original issue