원문정보
초록
영어
There is much research nowadays about artificial neural networks and artificial intelligence. In particular, research into cognitive models mimicking the human brain structure to simulate human consciousness are progressing at an active rate. This paper presents the cognitive model ‘HaeYum’, meaning ‘to think’ in Korean, based on Gerald Edelman’s Theory of neuronal group selection (TNGS) and Baar’s Global Workspace Theory (GWT). From these two theories regarding consciousness, the HaeYum model borrows the concepts of categorized memory and workspace. The HaeYum model explains consciousness as an operational mechanism dealing with categorized memory and represents it as a computational model, which could become a crucial theory regarding artificial intelligences that thinks like humans, or artificial general intelligence. This paper outlines the HaeYum model and explain its working mechanism of the memory-calling phase and the memory-saving phase.
목차
1. INTRODUCTION
Purpose of study
2. MATERIALS AND METHODS
Global Workspace Theory
Neural darwinism
3. RESULTS
Categorized memory
HaeYum Model
Memory calling state
Attention system
Oblivion
Consciousness occupation
Sensibility of sense
Exercise memory
Memory calling
Memory saving state
Conceptual categorization
Creation of memory
Memory separation
4. CONCLUSION AND DISCUSSION
5. References