Evaluation of the Notification Oriented Paradigm Applied to Sentient Computing

The Notification-Oriented Paradigm (NOP)defines a new approach for conception, structuring, and execution of software leading to performance improvements, organization of causal knowledge, and decoupling of programming entities when compared to standard solutions based upon current paradigms, such as the Imperative Paradigm (IP) and the Declarative Paradigm (DP). These paradigms present drawbacks such as waste of processing capacity and coupling among entities, due to their orientation to a monolithic inference mechanism that is based on causal evaluation implemented by means of searches over passive computational entities. The NOP, however, proposes a different approach of structuring software and realization of its inferences, which are based upon small, smart, and decoupled collaborative entities that interact by means of precise notifications. In this way, NOP achieves responsiveness, distributiveness, consistency, and robustness. These features are among the demands of Sentient Computing that combines ubiquitous and 'traditional' computing providing monitoring actuation on a given environment. This paper analyzes the NOP applicability to Sentient Computing by confronting the demands of the latter with the characteristics of the former. As conclusion, the NOP is considered applicable to this sort of computational system.