A Named Entity Recognition and Topic Modeling-based Solution for Locating and Better Assessment of Natural Disasters in Social Media

Over the last decade, similar to other application domains, social media content has been widely explored for disaster informatics. However, due to the unstructured nature of the data, several challenges are associated with disaster analysis in social media content. For example, social media content is generally very noisy containing disaster-related words (e.g., floods, wildfires) used in a different context. Similarly, most of social media posts either do not contain geo-location information or do not represent the actual disaster location. To fully explore the potential of social media content in disaster informatics, access to relevant content and the correct geo-location information is very critical. In this paper, we propose a three-step solution to tackling these challenges. Firstly, the proposed solution aims at the classification of social media posts into relevant and irrelevant posts followed by the automatic extraction of location information from the posts' text itself through Named Entity Recognition (NER) analysis. Finally, to quickly analyze the topics covered in large volumes of social media posts, we perform topic modeling resulting in a list of top keywords, that highlight the issues discussed in the tweet. For the classification of the tweets, we proposed a merit-based fusion framework combining the capabilities of multiple Large Language Models (LLMs), obtaining the highest F1-score of 0.933 on a benchmark dataset. For the Location Extraction from Twitter Text (LETT), we evaluated four LLMs obtaining the highest F1-score of 0.960. For topic modeling, we used the BERTopic library to discover the hidden topic patterns in the relevant tweets. The experimental results of all the components of the proposed end-to-end solution are encouraging and hint at the effectiveness of social media content and NLP in disaster management.