This study adopts a hybrid approach combining Systematic Literature Review (SLR) and Automated Content Analysis (ACA) to explore the application of Named Entity Recognition (NER) in disaster contexts. The growing demand for real-time, accurate information during disasters underscores the critical role of advanced information extraction techniques in enabling rapid decision-making and emergency response. This review investigates research trends, methodologies, and challenges in NER applications tailored for disaster management. Traditional rule-based NER methods often struggle with adaptability, whereas machine learning and deep learning approaches offer improved flexibility and scalability. Nonetheless, applying NER in disaster scenarios presents significant challenges, including handling complex linguistic variations, limited availability of annotated datasets, and processing diverse, unstructured text sources. By synthesizing current research, this study identifies key datasets, preprocessing methods, feature extraction techniques, and evaluation metrics that shape the development of NER models in disaster management. The findings aim to provide actionable insights into existing gaps and future directions, fostering the advancement of more robust and effective NER solutions to enhance disaster response and management worldwide.
Shidik et al. (Wed,) studied this question.