A Computational Model of the IAD framework

This repository implements a computational model of Elinor Ostrom's Institutional Analysis and Development framework. It includes the interpreter to the Action Situation Language (ASL) and the game engine to automatically generate extensive-form games from ASL descriptions.

Requirements

norms-games requires a working installation of the following:

• Python 3
• SWI-Prolog
• The PySwip package

Usage

To install the norms-games package, first clone the repository in your local file system:

git clone https://github.com/nmontesg/norms-games.git

Navigate to the root of the norms-games repository and install the package in editable mode using pip, from the environment you are using for your project:

cd /your/local/path/ngames
pip install --editable .

For the time being, the norms-games package requires to download a local copy of the source code (using git clone). The path to the package should then be appended to your Python path.

The examples directories has some illustrations on how to use the basic functions. Basically, you should create your ASL description in three distinct files:

• agents.pl
• states.pl
• rules.pl

Then, to construct the extensive-form game semantics of your description, it is enough to call:

build_full_game(<path_to_ASL_description>, <identifier>, (threshold=..., max_rounds=...))

See the documentation for further details.

References

Ostrom, E. (2005). Understanding Institutional Diversity. Princeton University Press.

Montes, N., Osman, N., & Sierra, C. (2021). Enabling Game-Theoretical Analysis of Social Rules. In Artificial Intelligence Research and Development (Vol. 339, pp. 90–99). IOS Press. https://doi.org/10.3233/FAIA210120

Montes, N., Osman, N., & Sierra, C. (2022). A Computational Model of Ostrom’s Institutional Analysis and Development Framework. Artificial Intelligence (Vol. 311). Elsevier. https://doi.org/10.1016/j.artint.2022.103756