The belief-structures model of Chopra and Parikh stands out as an exemplary framework adept at depicting both the statics and dynamics (revision) of beliefs held by agents with limited resources. In this article, following the research direction of our recent work (Aravanis, 2024), we further extend the belief-structures model in the realm of belief contraction. Furthermore, we present a natural method for adapting this new belief-structures model for contraction to a style of contraction à la Alchourrón, Gärdenfors and Makinson, retaining the favourable characteristics of belief structures. Under certain realistic assumptions, a collection of interesting properties of the proposed adaptation method is pointed out, by investigating the satisfiability and violation of principal postulates of the belief-revision theory. Last but not least, we investigate the interrelation between belief-structures contraction and revision, in the light of the well-known Levi and Harper Identities. The proposed approach could find application in contemporary areas like digital twins, where ensuring reliable and flexible belief changes is vital for precise and dynamic simulations that mirror real-world environments.
