A heterogeneous multi-attribute case retrieval method based on neutrosophic sets and TODIM for emergency situations

Area of Science:

• Decision Sciences
• Artificial Intelligence
• Information Science

Background:

• Effective emergency decision making (EDM) relies on retrieving relevant historical cases.
• Existing methods struggle with heterogeneous data formats in case retrieval.
• A robust method is needed to handle diverse attribute types for better EDM.

Purpose of the Study:

• To develop a heterogeneous multi-attribute case retrieval method for EDM.
• To incorporate five distinct attribute formats: crisp numbers, interval numbers, intuitionistic fuzzy numbers, single-valued neutrosophic numbers (SvNNs), and interval-valued neutrosophic numbers (IvNNs).
• To generate emergency alternatives by identifying the most suitable historical case.

Main Methods:

• Proposed a similarity measurement for interval-valued neutrosophic numbers (IvNNs).
• Calculated attribute similarities across five data formats and determined attribute weights using an optimal model.
• Computed case similarities, constructed a set of similar historical cases, and applied the TODIM method for utility evaluation.

Main Results:

• Successfully retrieved similar historical cases based on heterogeneous attributes.
• Determined the most suitable historical case using combined similarity and utility evaluations.
• Generated emergency alternatives, demonstrating the method's efficacy through a case study.

Conclusions:

• The proposed heterogeneous multi-attribute case retrieval method is effective for EDM.
• The method outperforms existing approaches in validity and superiority.
• This approach enhances the generation of emergency alternatives by leveraging diverse historical case data.