An integrated machine learning approach for identifying emergency rescue messages on social media during natural disasters

Abstract During large-scale disasters, emergency call centers are often overwhelmed by the large volume of rescue requests and calls for help. Consequently, people are turning to social media platforms to seek assistance. Rescue information posted on these platforms is extremely valuable for first responders to make informed rescue decisions. Therefore, the automatic identification of these requests from the vast amount of data posted on social media during crises is critical yet challenging. This work presents our ongoing research on applying deep learning techniques to extract actionable rescue information from social media during crises. We proposed a novel deep learning model that integrates a fine-tuned BERT to extract low-level statistical features and rule-based Regex filters to extract problem-specific features for emergency tweet identification. The proposed model was evaluated on labeled tweets collected from three hurricane events (Harvey, Ian, and Ida). Experimental results showed that our model performed better than several machine learning and deep learning methods in terms of the Area Under the Precision-Recall Curve (AUC-PR) metric for all events. This study contributed to the crisis informatics literature by introducing a novel deep learning approach for automatically identifying actionable information in social media, which can be adapted for similar natural language processing (NLP) tasks.