Adaptive Information Sharing with Ontological Relevance Computation for Decentralized Self-Organization Systems

Area of Science:

• * Artificial Intelligence
• * Swarm Intelligence
• * Complex Systems

Background:

• * Self-organizing systems exhibit decentralized control, relying on local coordination among simple agents.
• * Effective adaptation to environmental changes in these systems often requires inter-agent communication and knowledge sharing.
• * Lack of global state observation necessitates robust information exchange mechanisms for cooperative behavior.

Purpose of the Study:

• * To propose and evaluate a novel information sharing mechanism for decentralized self-organizing systems.
• * To enhance individual member adaptation and collective system optimization through improved information exchange.
• * To design an information sharing process analogous to human mechanisms, leveraging semantic relationships.

Main Methods:

• * Development of an independent decision phase for information sharing within self-organizing systems.
• * Utilizing ontology graphs to evaluate semantic relationships between information pieces and local knowledge.
• * Implementing a mechanism where relevant information updates local knowledge, reinforcing precise information sharing.

Main Results:

• * Simulations demonstrate efficient information sharing capabilities of the proposed mechanism.
• * Experimental results confirm the system's ability to achieve optimal adaptive self-organization.
• * The semantic evaluation approach leads to more relevant information acquisition and model reinforcement.

Conclusions:

• * The proposed information sharing mechanism significantly improves the adaptability of decentralized self-organizing systems.
• * Leveraging semantic relationships and ontology graphs is an effective strategy for enhancing knowledge exchange.
• * This approach facilitates more precise and efficient information sharing, leading to optimized system performance.