A computational framework for behaviour adaptation: the case for agents and business processes

Behaviour adaptation is a critical desideratum in almost all kinds of computational machinery. The challenge is to address the limitations of pre-programmed behaviour, which cannot anticipate the potentially vast range of possible situations in which that behaviour would need to be deployed or the potential adversarial behaviour of other entities in the operating context. This dissertation seeks to address this challenge in the context of two currently popular classes of computational machinery: intelligent agents and business processes.