본문 바로가기
  • Home

KEN Computational framework based Chain_of_Thought Reasoning

  • Journal of Software Forensics
  • Abbr : JSF
  • 2026, 22(2), pp.155~163
  • DOI : 10.29056/jsf.2026.06.14
  • Publisher : Korea Software Assessment and Valuation Society
  • Research Area : Engineering > Computer Science
  • Received : June 2, 2026
  • Accepted : June 20, 2026
  • Published : June 30, 2026

Jeong Yon Shim 1

1강남대학교

Accredited

ABSTRACT

Despite rapid advancements in artificial intelligence, existing symbolic and non-symbolic automated reasoning engines typically assume an emotionally neutral environment, failing to reflect emotional contexts critical to human cognition. To design a human-friendly and sophisticated AI system, this study proposes the Knowledge Emotion Network (KEN), a computational framework that structurally integrates knowledge structures with emotional elements. The proposed model employs the 3D VAD (Valence-Arousal-Dominance) emotional vector to dynamically adjust the activation thresholds of knowledge nodes and the inference weights of edges, thereby generating situation-sensitive Chains of Thought (CoT). Simulation results demonstrated that positive emotional states lead to deeper and broader exploratory reasoning paths, whereas negative emotions restrict inference to initial stages as a form of risk-aversion behavior. By structurally bridging affective computing and knowledge-based reasoning, this framework is expected to serve as a core intelligence engine for future human-computer interaction (HCI) agents and personalized AI counseling systems.

Citation status

* References for papers published after 2024 are currently being built.