Details

Published on Monday, 22 January 2024 11:40

Over the years, several efforts have informed standards for the sharing of computational models. Notable efforts include the “Overview, Design concepts and Details protocol”; the creation of the Network for Computational Modeling in the Social and Ecological Sciences (CoMSES); and “Transparent and comprehensive ecological modeling.” However, there are no common standards nor clear processes for sharing computational models at the point of publication. This complicates the task of effectively planning and resourcing research projects that intend to promote open and ethical modeling efforts across scientific domains.

The ModelShare Program is designed to respond to the above challenge. Through a series of interactive, online and free workshops, we are bringing the scientific community together to inform the development of guidelines for researchers to effectively implement model-sharing practices throughout the modeling lifecycle.

The workshops will take place between mid-February and late May 2024, and enable dialogue amongst data scientists, academic publishers, computational modelers, and the wider scientific community. Each workshop will host thought-provoking presentations by – or panel discussions with – leading experts. Meanwhile, the ModelShare team will ensure that events are welcoming and inclusive of diverse experiences, adhering to the Open Modeling Foundation’s (OMF) Code of Conduct.

Whilst the discussions will be community-driven, some questions we envisage tackling are:

• Should modeling standards be domain-agnostic or -specific?

• How do the FAIR and CARE principles apply to model-sharing?

• What support do publishers need to preserve, share, and cite  open models?

What’s more, the workshops will be steered by overarching topics:

• Guidance and Standards: Approaches and practices to describing and documenting models;

• Development Environments and Platforms: Model development in software development platforms and environments (e.g., GitHub);

• Research Computing: Model development and testing, data analytics, and workflows;

• Preservation and Publication: Model preservation in repositories and sharing in publications;

• Publication and Discoverability: Model metadata and availability in academic databases and search indexes;

• Sustainability: Machine learning, computational models, and ethics; and

• Assessment: Reflection on open science and FAIR indicators for models.

And the first workshop has been announced! On February 20th, at 15:00 (UTC), Volker Grimm (Open Modeling Foundation, Helmholtz Centre for Environmental Research – UFZ, and University of Potsdam) will deliver a presentation called “Developing standards for modeling is critical and exciting: lessons from ODD and TRACE.” To ensure you make it to the workshop,  join the ModelShare Google Group for email updates about the program.

The ModelShare Program is led by the OMF’s Certification Working Group,  enabled in part by generous support from the Digital Research Alliance of Canada and the Alfred P. Sloan Foundation, and coordinated by Kairoi.

Open Modelling Foundation project, we share the vision to enable open and ethical modeling efforts across domains of human and Earth systems sciences, mathematics and engineering. We strive to work together to provide an integrated representation of the complex world in which we live. And we hope you will be a part of this exciting journey.