Abstract
The article deals with the idea of integrating the concept of agent-based modeling of social processes and principles of psycological analysis of individual and collective ac-tions. Agent-based approach to computer simulation of complex processes in the eco-nomics, history, sociology and political science does not have a unified methodological basis neither in terms of the institutional structure of society, nor from the point of indi-vidual behaviour. Agents’ architecture in most of the models is based on the principles of so-called “folk psycology”, which is obviously not adequate to complex social sys-tems’ analysis. We suggest to model agent’s decision making in concordance with psycological studies of social behaviour. The start point of our research is the concept of an American psychologist Clark Hull, who gave a scientific explanation of common laws of behavior and how they can be replicated by a computer program. Hull’s model shows the relation between psycological needs and social reactions in terms of a state machine. We use Hull’s model to reproduce individual behaviour in the social environment. We assume that the agent has a unified estimate of the environment that he wants to maximise, and this motivation is his basic need. Another assumption is that the agent has an image of a better social environment, be it a different place or a political regime, and a potential opportunity to choose between them. Any changes require efforts and costs, that’s why the agent does not react immediately. Agent’s natural inertia is taken into concideration in the estimate called “the level of accumulated dissatisfaction”. All the estimates are normalised to the range [0; 1], which makes them easily computable. We show how the proposed estimates might be used for modeling various socially significant decisions: education, employment, migration, social tension. The logics of the agent’s decision making process is reproduced in terms of the finite state machine, based on the dynamically changing ratio of the environmental estimates, similarly to Hull’s model.
Comments
No posts found