Registration

Organisers: Rok Novak, Zuzanna Kurowska, Deniz Sirin
Duration: 3 hours

Communicating social simulation research poses specific challenges, from explaining model structure and assumptions to conveying uncertainty and relevance to non-expert audiences. This 3-hour workshop addresses science communication as an integral part of social simulation practice. Building on the goals of a proposed  Special Interest Group within the European Social Simulation Association, the workshop combines shared exploration with emerging guidance.

The session will include short inputs on communication techniques and approaches that the organisers are actively developing and testing in their own work, alongside structured discussion and hands-on activities. Participants are invited to contribute their own cases, experiences, and communication challenges. Rather than presenting fixed best practices, the workshop aims to jointly examine what works, in which contexts, and why.

The goal is to collectively identify promising strategies, recurring pitfalls, and open questions, and to lay the groundwork for a more systematic approach to science communication within the social simulation community.

Organisers: Loïs Vanhée, Melania Borit, Bianca Luque Capellas
Duration: 3 hours

Models of human decision (MOODs) are a central component of agent-based social simulation (ABSS) that raises a variety of open ended questions investigated by the ABSS community as well as by  connected ones (e.g., affective computing, intelligent virtual agents). While MOODs design and integration within ABSS are sometimes said to intersect “art and science”, few arenas are available for discussing practical insights based on experience or joint interests and aspects on which members of our community could collaborate.

This workshop is meant to create such an arena. After a brief introduction on recent developments in the science and community tied to MOOD (the SIG-MOOD ESSA special interest group), the workshop will be organized as a networking space for facilitating exchanges among participants on concrete MOODs content (what do you model), methods (how do you model), and applications (what do you model for). At the end of the workshop, participants should be able to identify who their research relates to in the community, as well as collective interests and needs they can contribute to.

Organisers: Bruce Edmonds, Cezara Pastrav, Sofia Karlsson, Frank Dignum
Duration: 3 hours

There has been a long standing workshop and track on Qual2Quant where people discuss the use of qualitative data for Social Simulations. In this workshop we want to build on this and look specifically how ethnographic methods can be used to develop social simulations. As ethnographic methods emphasize the description of phenomena from the perspective of the humans that are involved, they provide many handles for modelling agents, processes, mechanisms and rules, as well as environmental contextual information that can be extremely relevant for the simulations of these phenomena. Through our work over the last few years, we have gained experience with modelling a simulation to support the management of mining disasters. But also on how teams adapt to disruptive innovation technology when they manage leakages on an extensive pipe networks carrying warm water for heating. And another on how people make decisions on energy consumption, which will be used to simulate the energy consumption of a new sustainable neighbourhood. Very different applications but similar methods were used.

In the workshop we plan to concretely explore how this type of ethnographic data can be used in a more systematic way to create social simulations.

Organisers: Kevin Chapuis, Flavien Loup—Hadamard
Duration: 6 hours
Requirements: Laptop with capacity to install GAMA

GAMA is an easy-to-use open-source modelling and simulation environment for creating spatially explicit agent-based simulations. It has been developed to be used in any application domain (e.g. urban mobility & planning, climate change adaptation, epidemiology, disaster evacuation strategy design) and with any ABM approach, including participatory modelling and data-intensive simulation.

The generality of the agent-based approach advocated by GAMA is accompanied by a high degree of openness, which is manifested, for example, in the development of plugins designed to meet specific needs or by the possibility of calling GAMA from other software or languages (such as R or Python). This openness allows the more than 2000 users of GAMA to use it for a wide variety of purposes: scientific simulation, scenario exploration and visualization, negotiation support, serious games, mediation or communication tools, the possibilities are endless!

We propose to organise two workshops. The first one to introduce the platform, the dedicated language called Gaml, and its resources (online documentation, tutorials, and more) to help you start your project powered by Gama.

The second one, more advanced, introduces how to use GIS data for simulation in GAMA. During this 3-hour workshop, a model of urban mobility & planning will support the training based on Gaml reusable building blocks.

Organisers: Aditya Khanna, Jonathan Ozik
Duration: 3 hours
Requirements: Familiarity with logistic regression, experience fitting logistic model in R will help.

Agent-based models increasingly rely on networked synthetic populations derived from multisource empirical data. Yet the process by which empirical summaries (e.g., mixing structures, degree distributions, geocoded location data) are translated into simulated networks is often ad hoc. The opacity and heuristic nature of the underlying workflows can be prohibitively intimidating for new users and difficult to reproduce even for experienced practitioners. This workshop focuses on workflow design, diagnostics, and judgment in generating networked populations using exponential random graph models (ERGMs) for social simulation. The underlying example will focus on modelling syringe-sharing networks to simulate vaccine interventions in an agent-based modelling framework. Using ERGMs, participants will work through key stages of a reproducible pipeline: defining network targets from empirical data, including geographic attributes; stepwise ERGM fitting; diagnosing failure modes (e.g., model convergence); and assessing alignment between simulated and target networks. The workshop emphasizes reproducibility, transparency, and explicit modelling choices, aiming to develop shared standards of practice that integrate both the art and technical science of network modelling.

The workshop comprises 6 modules of about 30 minutes each.

Organisers: Valerii Chirkov, Wataru Toyokawa
Duration: 3 hours
Requirements: Basic Python

This three-hour workshop introduces participants to simulation-based inference (SBI) and its applications to complex social models, bridging the gap between agent-based simulations and experimental research [2, 3]. The session is structured around interactive Python Jupyter notebooks and focuses on a practical, hands-on learning experience. Participants execute code and complete exercises immediately following each content section. The schedule is divided into two parts. First, we will cover the fundamental SBI workflow and best practices using the sbi Python package [1, 3], guiding attendees through toy inference problems (e.g., projectile motion). Second, we will introduce social learning models based on [5] and apply SBI to infer specific parameters (e.g., learning rate and social weight) from behavioural measures.

Participants are encouraged to bring a laptop and are required to have a basic Python 3 knowledge to fully engage with the coding activities.

References

[1] Boelts, J., Deistler, M., Gloeckler, M., Tejero-Cantero, Á., Lueckmann, J.-M., Moss, G., Steinbach, P., Moreau, T., Muratore, F., Linhart, J., Durkan, C., Vetter, J., Miller, B. K., Herold, M., Ziaeemehr, A., Pals, M., Gruner, T., Bischoff, S., Krouglova, N., … Macke, J. H. (2025). sbi reloaded: A toolkit for simulation-based inference workflows. Journal of Open Source Software, 10(108), 7754.

[2] Cranmer, K., Brehmer, J., & Louppe, G. (2020). The frontier of simulation-based inference. Proceedings of the National Academy of Sciences, 117(48), 30055–30062.

[3] Deistler, M., Boelts, J., Steinbach, P., Moss, G., Moreau, T., Gloeckler, M., Rodrigues, P. L. C., Linhart, J., Lappalainen, J. K., Miller, B. K., Gonçalves, P. J., Lueckmann, J.-M., Schröder, C., & Macke, J. H. (2025). Simulation-Based Inference: A Practical Guide (No. arXiv:2508.12939). arXiv.

[4] Ramalho, L. (2022). Fluent Python: Clear, concise, and effective programming (2nd ed.). O'Reilly Media.

[5] Toyokawa, W., Whalen, A., & Laland, K. N. (2019). Social learning strategies regulate the wisdom and madness of interactive crowds. Nature Human Behaviour, 3(2), 183–193.

Important Note: The option in the registration form is audience member for this workshop, not a participant. Participation is through the ESSA@Work programme, advertised separately. The ESSA conference provides a venue for this programme, with both a morning and afternoon session.

Organisers: Vivek Nallur, Katharina Luckner, Aytalina Kulichkina, Liu Yang, Samuel Ugo Ringier
Duration: 2 x 3 hours, full day (allows 8-10 participants to present their work-in-progress and receive feedback)

ESSA@work has a long tradition within the European Social Simulation community. It is based on the desire to give and receive feedback on work-in-progress (agent-based) models. Participants present a model they are working on to gather feedback and suggestions to improve, adapt and/or extend their model.

Participants present a model they are working on to gather feedback and suggestions to improve, adapt and/or extend their model. The model can be at any stage of development, however, it should have been at least partially explored by authors, and they should prepare clear and concise questions/problems that they are struggling with in relation to the model. Participants are asked to submit these questions ahead of time to allow for the preparation of feedback. Feedback to participants comes from two different sources: two expert modellers and the audience. The audience is invited to serve as ad-hoc experts by joining the expert panel in a fish-bowl set-up. In the feedback process, emphasis is placed on constructive exploration of possible solutions to the problems raised by the participants.