Proceedings of the Annual Meeting of the Cognitive Science Society

The development of self-adaptive systems has attracted lots ofattention as they can adapt themselves autonomously to en-vironmental dynamics and maintain user satisfaction. How-ever, there are still tremendous challenges remained. One ma-jor challenge is to guarantee the reusability of the system andextend the adaptability with the changing deployment environ-ments. Another challenge is to ensure the adaptability copingwith the open and complex environments with the existence ofunknown. To solve these problems, we introduce a concep-tual self-adaptive model, decoupling the environment with thesystem. This model is a two-layer structure, based on internalcauses and external causes from attribution theory. The firstlayer, determining how the internal causes affect the adapta-tion behaviors, is independently designed and reusable; whilethe second layer, mapping the relationship between externalcauses with internal causes, is replaceable and dynamicallybound to different deployment environments.