A Delphi-based approach to developing expert systems with the cooperation of multiple experts

Area of Science:

• Artificial Intelligence
• Knowledge Engineering
• Expert Systems

Background:

• Knowledge acquisition is a key challenge in developing knowledge-based and expert systems.
• Existing methods often focus on single expert knowledge, neglecting the benefits of multiple perspectives.
• The temporal dynamics of expert knowledge significantly impact inference accuracy, necessitating time-sensitive elicitation.

Purpose of the Study:

• To propose a novel approach for eliciting and integrating knowledge from multiple experts.
• To address the challenge of incorporating the time factor in expert knowledge acquisition.
• To enhance the development of effective expert systems through collective intelligence.

Main Methods:

• Development of a Delphi-based approach for structured knowledge elicitation from multiple domain experts.
• Integration of temporal considerations into the knowledge acquisition process.
• Validation of the approach through a case study on Severe Acute Respiratory Syndrome diagnosis.

Main Results:

• The proposed Delphi-based approach facilitates the integration of knowledge from multiple experts.
• The method demonstrates effectiveness in capturing time-dependent expert knowledge.
• The application in Severe Acute Respiratory Syndrome diagnosis highlights the approach's superiority over traditional methods.

Conclusions:

• The Delphi-based approach offers a robust solution for multi-expert knowledge acquisition in expert systems.
• Incorporating the time factor enhances the accuracy and reliability of expert system inferences.
• This methodology advances the field of knowledge engineering and expert system development.