본문 바로가기
  • Home

An Extraction Mechanism of Perception connected Semantic/Episodic Memory

  • Journal of Software Assessment and Valuation
  • Abbr : JSAV
  • 2024, 20(1), pp.65-73
  • Publisher : Korea Software Assessment and Valuation Society
  • Research Area : Engineering > Computer Science
  • Received : March 2, 2024
  • Accepted : March 20, 2024
  • Published : March 31, 2024

Jeong Yon Shim 1

1강남대학교

Accredited

ABSTRACT

These days, Chat GPT has already gone beyond simple knowledge processing and is replacing many parts of human intelligence and its functions are evolving at a rapid pace. The basis of chatbot research is human intelligence, which has adapted to harsh environments that has made survival possible for millions of years. In this study, human memory and its operating principles was examined from the perspective of human intelligence research and system modeling. So, we designed structured memory modeling to enable learning/recognition and understanding of objects at the same time. Based on short-term /long-term memory structures, we proposed a mechanism that can extract related knowledge & facts by directly connecting the learning/perception module to sementic/episodic memory. Through experiments, we tested extraction mechanism of the proposed system

Citation status

* References for papers published after 2023 are currently being built.