Evolving agents: communication and cognition

  • Authors:
  • Leonid Perlovsky

  • Affiliations:
  • Air Force Research Lab., MA

  • Venue:
  • AIS-ADM 2005 Proceedings of the 2005 international conference on Autonomous Intelligent Systems: agents and Data Mining
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Computer programming of complex systems is a time consuming effort. Results are often brittle and inflexible. Evolving, self-learning flexible multi-agent systems remain a distant goal. This paper analyzes difficulties toward developing evolving systems and proposes new solutions. The new solutions are inspired by our knowledge of the human mind. The mind develops language and cognitive abilities jointly. Real-time sensor signals and language signals are integrated seamlessly, before signals are understood, at pre-conceptual level. Learning of conceptual contents of the surrounding world depends on language and vice versa. This ability for integrated communication and cognition is a foundation for evolving systems. The paper describes a mathematical technique for such integration: fuzzy dynamic logic and dual cognitive-language models. We briefly discuss relationships between the proposed mathematical technique, working of the mind, applications to understanding-based search engines and evolving multi-agent systems.