Home // SIMUL 2021, The Thirteenth International Conference on Advances in System Simulation // View article

A Framework to Specify Agent-Based Models Using ODD* Protocol

Authors:
Cédric Grueau
Thyago Romagna Bendo
Alan Gavioli
João Junior Araújo

Keywords: ODD Protocol; ODD+D; ODD+2D; Linked Data; Agent-Based modelling.

Abstract:
The use of Agent-Based Models (ABM) is a popular approach to develop simulations in fields such as social sciences, geography and natural sciences. However, due to the complexity of the models, the sharing of models’ specification becomes an arduous process, which makes it difficult to validate and replicate this type of model. To assist in this process, experts from the domain have developed the ODD (Overview, Design concepts, Details) protocol to help systematize ABM descriptions. However, because it is a protocol, the latitude allowed in the description of the models and the format in which these descriptions are made only address part of the problem. On the one hand, it may remain difficult to understand the description, and on the other hand, the access and processing of the ODD descriptions are still a manual task. To solve the problem, a framework, based on an ontology, was developed to allow traceability of ABM described with ODD. Concepts from other metadata initiatives, such as Dublin Core and Codemeta, were included to add important data to the ODD protocol. An application was developed to facilitate the access to the models, which allows the analysis of the data source of the model and the extraction of the protocol components. An API was also developed that gives access to ABM descriptions and promotes the processing of ABM descriptions. The application is ready for a series of tests to validate our approach, the usability of the application and the utility of the framework.

Pages: 95 to 98

Copyright: Copyright (c) IARIA, 2021

Publication date: October 3, 2021

Published in: conference

ISSN: 2308-4537

ISBN: 978-1-61208-898-3

Location: Barcelona, Spain

Dates: from October 3, 2021 to October 7, 2021