Manuscript Title:

ADAPTIVE DECISION SUPPORT SYSTEMS USING GPT-4 FOR CRISIS MANAGEMENT

Author:

Dr. M HAFIZ YUSOFF, Dr. JULAILY AIDA JUSOH, Dr. TS. FATMA SUSILAWATI MOHAMAD, Dr. WAN MOHD AMIR FAZAMIN WAN HAMZAH, Dr. SYARILLA IRYANI AHMAD SAANY, Dr. YC ONG CHUAN

DOI Number:

DOI:10.5281/zenodo.12515689

Published : 2024-06-23

About the author(s)

1. Dr. M HAFIZ YUSOFF - Associate Professor, Dato, Deputy Vice Chancellor for Student Affairs, UniSZA, Malaysia.
2. Dr. JULAILY AIDA JUSOH - Faculty of Informatics and Computing, UniSZA, Malaysia.
3. Dr. TS. FATMA SUSILAWATI MOHAMAD - Associate Professor, Faculty of Informatics and Computing, UniSZA, Malaysia.
4. Dr. WAN MOHD AMIR FAZAMIN WAN HAMZAH - Faculty of Informatics and Computing, (UniSZA), Malaysia.
5. Dr. SYARILLA IRYANI AHMAD SAANY - Associate Professor, Faculty of Informatics and Computing, UniSZA, Malaysia.
6. Dr. YC ONG CHUAN - Faculty of Informatics and Computing, UniSZA, Malaysia.

Full Text : PDF

Abstract

Introduction: The increasing frequency and complexity of crises pose significant challenges to crisis management systems. Traditional decision support systems often struggle to adapt dynamically to the evolving nature of crises. This research explores the integration of GPT-4, a state-of-the-art natural language processing model, into adaptive decision support systems for crisis management. Problem Statement: Current crisis management systems are limited in their ability to provide real-time, contextaware decision support. The lack of adaptability hampers their effectiveness in addressing rapidly changing crisis scenarios. This research addresses the critical need for more responsive and flexible decision support systems in crisis management. Objective: The primary objective of this research is to design and implement an Adaptive Decision Support System (ADSS) using GPT-4 to enhance crisis management capabilities. The system aims to provide timely, contextually relevant information and recommendations to decision-makers during crises. Methodology: The study employs a multi-faceted methodology, combining literature review, system design, and empirical evaluation. The design phase involves the integration of GPT-4 into the decision support system architecture, enabling the system to adapt to dynamic crisis situations. The empirical evaluation assesses the system's performance in simulated crisis scenarios, measuring response time, accuracy, and user satisfaction. Results: Preliminary results indicate that the incorporation of GPT-4 significantly improves the adaptability of the decision support system, leading to more effective crisis management. The system's ability to process vast amounts of information in real-time contributes to better-informed decision-making during crises. The empirical evaluation reveals positive outcomes in terms of response time, accuracy, and user satisfaction. Conclusion: This research demonstrates the feasibility and efficacy of leveraging GPT-4 in creating adaptive decision support systems for crisis management. The integration of advanced natural language processing capabilities enables the system to adapt dynamically to evolving crisis scenarios, offering valuable insights and recommendations to decision-makers. The findings underscore the potential of GPT-4 in enhancing crisis management capabilities and suggest avenues for further research and implementation.


Keywords

Adaptive Decision Support Systems, GPT-4, Crisis Management, Natural Language Processing, Context-Aware Decision Making, Empirical Evaluation.