Developing best practices for an open library of archaeological ABM modules: lessons learned from other initiatives
Presentation for the session 367 "Agent-based Modelling of Socio-ecological Systems in Archaeology. Towards a New Research Community" at the EAA 2022 “[Re]integration” conference in Budapest (31 August - 3 September 2022).
EAA2022 page: https://www.e-a-a.org/eaa2022
Prepared with MS Microsoft PowerPoint
Authors
Angourakis, Andreas (Institute for Archaeological Studies, Ruhr University Bochum; McDonald Institute for Archaeological Research, University of Cambridge) - Riede, Felix (School of Culture and Society, Aarhus University) - Romanowska, Iza (Aarhus Institute of Advanced Studies, Aarhus University) - Verhagen, Philip (Vrije Universiteit Amsterdam) - Saqalli, Mehdi (GEODE, Université Toulouse II - Jean Jaurès) - Taelman, Devi (Department of History, Archaeology, Arts, Philosophy and Ethics, Vrije Universiteit Brussel) - Vlach, Marek (Institute of Archaeology of the Czech Academy of Sciences) - Sikk, Kaarel (Masaryk University) - Galán Ordax, José Manuel (Universidad de Burgos) - Brughmans, Tom (School of Culture and Society, Aarhus University)
Abstract
The main objective of the Network for Agent-based modelling of Socio-ecological Systems in Archaeology (NASSA) is to create and manage an open library of agent-based modelling (ABM) modules (not entire models) to be used by our community and beyond. The library will hold contributions from a diverse community of archaeological modellers as well as non-archaeologists in the form of reusable code modules that can be recycled and repurposed in new models. This open repository will improve communication, reproducibility and reusability helping to streamline model development and facilitate skilled and specialist knowledge acquisition for learners. The added value of the repository will be generated by the explicit use of transparent standards, strong metadata and active feedback from users. Although singular in its specific domain, this is neither the first nor the only initiative that attempts to openly collate and curate collective knowledge.
This paper will focus on identifying a non-exhaustive list of “good” practices and illustrate them with real-world examples taken from our experience with other similar initiatives (CoMSES Computational Model Library, R and Python packages, specialised modelling systems or libraries, large projects in archaeology with strong ABM components, and open databases). We bring to attention the lessons learned in other disciplines that address socio-ecological systems, such as earth sciences, ecology, sociology and economics, but also reflect on what are the constraints and needs of modelling for archaeological questions with archaeological data. Is archaeology a special case, and if so, what are the prerequisites of the field for making this a successful endeavour? Against the canvas of experiences collected, we raise awareness about both the design and maintenance of the library itself, as well as the preparation and submission of modules by the broader community.