Cognitive architectures are in general used to create intelligent systems that exhibit behaviors similar to those of humans. The creation of such structures is based on diverse approaches intended to explain the mind, whose attempt to describe the “cognitive” processes is made through embodied perception-action theories. These processes are denoted as the different abilities of human beings to act in everyday life situations and events, in particular, the actions guided by motivations and emotions, plans and decisions, in order achieve well established or new goals.
To study cognitive processes, these are abstracted at three distinct levels:
- The components that made up the cognitive process.
- The emerging interaction between components.
- The interaction between sets of cognitive processes that made up a whole cognitive system.
Based on this, existing cognitive architectures propose the use of approaches mentioned above to human mind to develop complex and intelligent systems. Moreover, they attempt to establish a series of requirements and characteristics that cognitive architectures must have. Then, they will be likely to be used for more “believable” intelligent agents, or complex systems that aid researchers on their studies of the mind. Nevertheless, most of the cognitive architectures suffer a serious issue of knowledge representation, and consequently, this limits cognitive systems.
Thus, the scope of the present research work is the creation of a Cognitive Architecture based on the brain areas and their functions taken from neurosciences' results. Our approach's assumption is that Neuroscience and Neurophysiology are the fundamental theories used to address the knowledge representation issue. In particular, biological foundations are used to explain the construction of cognitive systems based on components interaction.
Currently, we are working in the development of two main systems: