Analogue driven handling of emergent complexity

Abstract

Complexity Science has noted as the next evolution of the science in which the main idea is a global emergent behavior, which is resulting from the interaction between its local, dynamic, complex, interconnected, distributed, and uncertain entities. In Computer Science; modelling of complex systems are not naive and many technologies have been used for this specially due to the advancements in Artificial Intelligence. This paper presents a novel approach to model complex systems through relations on analogy making. Analogy making is considered as a mapping of knowledge between two domains. Multi Agent Technology has been used to facilitate the underneath support for this, while primarily using the power of relating in different domains as the working mechanism for agents. Agents are used to isolate an appropriate domain in which the same problem that happens to be analogically feasible; and for this domain selection, communication and negotiation features of agent technology have been used. The model has been partially scrutinized in the geometric analogy domain for empirical validations and the idea is still in a conceptual level.