Worlds as a unifying element of knowledge representation
Loading...
Authors
Scally, Jonathan Richard
Issue Date
2014-08
Type
Electronic thesis
Thesis
Thesis
Language
ENG
Keywords
Cognitive science
Alternative Title
Abstract
Cognitive systems with human-level intelligence must display a wide range of abilities, including reasoning about the beliefs of others, reasoning about past and future versions of themselves, consideration of hypothetical and future situations, and reasoning with uncertainty, counterfactuals, quantified objects, and abstract situations. While each of these deals in some way with reasoning about alternative states of reality, no single knowledge representation framework deals with them in a unified, general, and scalable manner. As a consequence it is difficult to build cognitive systems for domains that require each of these abilities to be used together. To enable this integration, this thesis proposes a representational framework based on synchronizing an agent's knowledge between different "worlds". Using this framework, each of these tasks can be reformulated into a reasoning problem involving worlds. This demonstrates that by building upon small set of fundamental principles, cognitive systems can achieve representational parsimony, deep integration, and broad new abilities. The thesis also analyzes a working implementation of the representational framework, built upon the Polyscheme cognitive architecture, and discusses lessons for future worlds-based cognitive architecture development.
Description
August 2014
School of Humanities, Arts, and Social Sciences
School of Humanities, Arts, and Social Sciences
Full Citation
Publisher
Rensselaer Polytechnic Institute, Troy, NY