A Taxonomy of Computation and Information Architecture

Paper in proceeding, 2015

Nowadays computation is typically understood through the Turing machine model, in the fields of computability, computational complexity and even as a basis for present-day computer hardware and software architectures. Those are technologies designed in the first place to process data. Being description of data manipulation, Turing model of computation presents only one aspect of computation in the real world – an abstraction of the execution of an algorithm. However, several other possible aspects of computation, even those existing in today’s applications, are left outside, thus adequate models in distributed, self-organized, resource-aware, adaptive, learning computation systems are needed. This paper presents taxonomy of existing models of computation. It is connected to more general notion of natural computation, intrinsic to physical systems, and particularly cognitive computation in cognitive systems. We see Turing model of computation as a basic mechanism which can be used to build more complex computational architectures, that in combination with interaction with the environment (learning) give advanced information-processing behaviors in cognitive systems.