Date of Award

5-2017

Type

Thesis

Major

Master of Science

Department

TSYS School of Computer Science

Abstract

Learning is the modification of a behavioral tendency by experience. Memory and reasoning are the most important aspects for learning in humans; information is temporarily stored in the short-term memory and processed, compared with existing memories and stored in long-term memory, and can be re-used when needed. One way to describe an organized pattern of thought or behavior and the categories of information along with their relationships is by using schemas. A cognitive script is one form of a schema that evolves over multiple exposures to the same set of stimuli and/or repeated enactment of a particular behavior. This research aims to provide a comparative study between three cognitive systems/tools designed to allow learning, by using cognitive scripts representation. Since retrieving and re-using past experiences is the core of any learning process, the focus of this thesis is to examine the current existing cognitive systems and tools to evaluate their ability to retrieve past experiences. SOAR, myCBR and Pharaoh are three systems considered for this thesis. Linear and multi-branched cognitive scripts were considered in order to measure the capacity of those systems to allow learning using cognitive scripts representation. The results of this work show that SOAR, myCBR and Pharaoh took almost the same time to retrieve a set of similar cognitive scripts to a query script. However, SOAR was able to retrieve one similar script only, while myCBR and Pharaoh were able to retrieve multiple scripts. Pharaoh tops the other two system in its ability to handle multibranched scripts of different sizes and the way it considers context.

Share

COinS