[3I5-GS-11] AI and Society:

[3I5-GS-11-04] Predicting and simulating community activities using a digital twin that integrates logic models and local resident models

Probabilistic modelling for community regeneration

〇YOICHI MOTOMURA¹, Koji Kitamura¹, Kazuya Yamashita¹, Eiichi Sakurai¹, Megumi Inoue¹ (1. National Institute of Advanced Industrial Science and Technology)

Keywords:Digital Twins, Probabilistic models, Digital transformation, Local resident models

In order to solve social issues, we are developing technology to build and utilize digital twins that combine cyber and physical worlds. In particular, to address the issue of revitalizing local communities, where aging and population decline have made it difficult to sustain these activities, we will describe technology that collects logic models related to community activities and behavioral and survey data of local residents, and predicts and simulates community activities using digital twins that integrate logic models and resident models. We will show a method of performing probabilistic inference by integrating a Bayesian network, which handles the diverse values, non-determinism, and situation-dependence of residents' attitudes and actions, with a logic model that represents community activities.