Norm mining, identification, and detection: a systematic literature review.

Area of Science:

• Artificial Intelligence
• Multi-Agent Systems
• Socially Aware Computing

Background:

• Norms are essential for regulating behavior in multi-agent systems, promoting cooperation and conflict resolution.
• Understanding norm identification is key to developing sophisticated and reliable autonomous systems.
• Existing research presents diverse approaches to norm management within agent societies.

Purpose of the Study:

• To systematically review and categorize existing methods for norm identification in multi-agent systems.
• To evaluate the effectiveness of these methods, particularly in dynamic and uncertain environments.
• To identify current limitations and suggest future research directions.

Main Methods:

• Systematic literature review methodology.
• Analysis of 35 selected studies on norm identification.
• Categorization of methods for norm detection, synthesis, and adaptation.

Main Results:

• Identified and categorized various norm identification techniques.
• Assessed the performance of methods in simulated dynamic and uncertain environments.
• Highlighted significant gaps in scalability, adaptability, and real-world applicability of current approaches.

Conclusions:

• Current norm identification methods face challenges in complex, real-world scenarios.
• Future research should focus on integrating Large Language Models (LLMs) and interdisciplinary collaboration.
• Advancing socially aware autonomous systems requires more robust and adaptable norm identification strategies.