Enhancing the Learning Capability in SOAR Cognitive Architecture

Abstract

Intelligent agents enable the implementation of artificial intelligence by making decisions. The supporting structure for an intelligent system is defined by a cognitive architecture. In this paper, with the help of Soar cognitive architecture, we are trying to empower the advanced agent system by enhancing their intelligence using knowledge representation, environment, and capabilities. Soar’s Semantic memory stores the content of agents’ information and knowledge in the form of facts and its Episodic memory supports a collection of cognitive capabilities that help an agent observe its environment, think critically, be more capable, and therefore learn.

In our work, we are enhancing the two-digit arithmetic addition agent by increasing its number of digits i.e. from two to three, thereby making it more capable and using rote learning to make it intelligent.