A Triangular Similarity Measure for Case Retrieval in CBR and Its Application to an Agricultural Decision Support...

Area of Science:

• Artificial Intelligence
• Agricultural Science
• Decision Support Systems

Background:

• Case-based reasoning (CBR) aids decision-making through retrieve, reuse, revise, and retain steps.
• Effective case retrieval is crucial for CBR success, with existing angle-based and distance-based methods having limitations.
• Inaccurate case retrieval can occur with traditional similarity measures in extreme scenarios.

Purpose of the Study:

• To propose a novel triangular similarity measure for enhanced case retrieval in CBR.
• To address the limitations of angle-based and distance-based measures in identifying similar cases.
• To improve the accuracy and robustness of case retrieval in decision support systems.

Main Methods:

• A triangular similarity measure was developed to identify commonalities between cases.
• Case-based reasoning was applied to an agricultural decision support system for pest management.
• The proposed measure was tested using 300 new pest management cases.

Main Results:

• The triangular similarity measure achieved an average accuracy of 91.99% in retrieving the most similar case.
• The proposed measure demonstrated superior accuracy and robustness compared to existing methods.
• The system provided quick decision support for farmers managing pest problems.

Conclusions:

• The triangular similarity measure is effective for accurate and robust case retrieval in CBR.
• This approach significantly enhances decision support in agricultural pest management.
• The proposed method offers a valuable improvement over traditional similarity measures.