NFDI4DS is hosting a series of Mini Hackathons at ZB Med in Cologne.
6th - 8th November 2023: FAIRification Game for Software
Building on top of the FAIRification game for rare disease data, we want to create a FAIRification game for software based on the FAIR4RS principles.
In this mini-hackathon we will work on:
- Introduction to the FAIRification game for rare disease so newcomers can better contribute to the Software FAIRification game
- Core FAIRification game for software
- If time allows, we will start analyzing customization needed for rare disease and the machine learning use cases
21st - 24th November 2023: Metadata for Machine Learning
In collaboration with the RDA FAIR4ML IG we will discuss and define some concepts relevant for metadata Machine Learning.
In this mini-hackathon we will work on three topics:
- 21-22. Nov.2023: Machine Learning Lifecycle.
The ML lifecycle comprises finding data for the model, building, testing, validating, sharing, reusing. In this project, we want to define the ML lifecycle and identify its connections to FAIR principles. - 22-24. Nov.2023: Machine Learning Lifecycle - Visualization.
Once we have the ML lifecycle defined, we could work on an interactive visualization (e.g., D3 combined with Vue) that can later be connected with the NFDI4DS registry. In these two days, we could try and connect to some ML-related platforms so we can see some results. - 23-24. Nov.2023: Metadata for Machine Learning.
Based on ML Cards and Data Cards, we want to define a mapping to schema.org types and properties, including new elements as needed.
28th November - 1st December 2023: Machine-actionable Software Management Plans
There are different Software Management Plans available, e.g., ELIXIR and Max Planck Digital Library (MPDL), which are currently based on questions and answers. An initial effort on a machine-actionable layer on top of the ELIXIR SMP in the form of a metadata schema is already available but still needs better alignment with other efforts.
In this mini-hackathon we will work on:
- Alignment to the MPDL SMP, maybe e-Science Centre SMP, and the clinical case
- Metadata support in research management software, in particular the Research Data Management Organiser (RDMO), currently the de facto option in Germany
- Metadata extraction from GitHub and GitLab repositories and its integration into RDMO