Extracting Useful Emergency Information from Social Media: A Method Integrating Machine Learning and Rule-Based Classification

This study proposes a machine learning and rule-based integration method (MRIM) to extract useful emergency information (EI) from social media user-generated content (UGC). The MRIM outperforms pure machine learning and rule-based methods in classifying EI from microblog data about the July 2021 Zhengzhou rainstorm. The study's findings highlight the impact of microblog characteristics, such as word count, address and contact information, and user attention, on the performance of the MRIM. The research demonstrates the feasibility of integrating machine learning and rule-based methods for effective EI extraction and provides actionable suggestions for emergency information management.

Hongzhou Shen, Yue Ju, Zhijing Zhu