Host: The Japanese Society for Artificial Intelligence
Name : The 38th Annual Conference of the Japanese Society for Artificial Intelligence
Number : 38
Location : [in Japanese]
Date : May 28, 2024 - May 31, 2024
Multi-agent simulation has been used by computational social scientists to study complex social problems such as sustainability, creativity, etc. To effectively analyze the dynamics of such social problems, it is often necessary to represent the social phenomena of interest with a single static state or indicator, which may lead to an oversimplification of complex social problems. In this study, we propose to use model checking to overcome this problem. Model checking is a formal verification technique that can be used to verify that properties can be satisfied by traversing all possible states of the system. This approach has been successfully used in the design and development of hardware and software systems, and we propose to apply it to socio-dynamic systems to facilitate the modeling of complex social problems. First, based on the existing computational social models, we describe the social problems of interest by transition systems, where each state is defined by multiple indicators. Then, by defining the social properties to be verified using a formal language, the social properties can be verified by model checking tools to see if they are satisfiable under certain conditions. In this study, we use a multi-agent model based on opinion dynamics as an example to demonstrate the potential of the proposed approach.