How effective is twitter (X) social media data for urban flood management?

Abstract

Social media has been an instant source of information for natural disasters, such as urban floods, throughout the world. Ex-post evaluation of the information is considered more important after a flood disaster with devastating consequences on critical infrastructure, the environment, and the well-being of the society. Research proposes an evaluation framework for integrating the disorganized online public opinion on urban flood disaster events towards emotional and conceptual characteristics for better ex-post analysis and public participation. Social media posts on an urban flood were acquired using a search engine, and then sentiment analysis, topic modeling, and spatial–temporal analysis were performed to generate measures of online public opinion about the urban flood crisis. Twitter (X) is the most popular microblogging service presently available. As per the methodology, we analyzed all tweets regarding the 2022 urban flood in the cosmopolitan city (Karachi) of Pakistan from users all over the world. The evaluation results demonstrated that the distribution patterns of post intensities and emotion polarity in response to the floods emphasized crucial aspects with contradictory emotions and highlighted the strategic implications. Experimental results on real datasets show relatively better performance than the baseline and state-of-the-art approaches and achieved the highest 91% score. Online public opinion is a valuable supplement to ex post-disaster evaluation as it helps the project (e.g., flood management) better perform and provides suggestions for future flood mitigation, especially public participation management.