Social Media Information Classification of Earthquake Disasters Based on BERT Transfer Learning Model

被引：1

作者：

Lin, Sen ^{[1
]}

Liu, Beibei ^{[1
]}

Li, Jianwen ^{[1
]}

Liu, Xu ^{[1
]}

Qin, Kun ^{[2
]}

Guo, Guizhen ^{[1
]}

机构：

[1] National Disaster Reduction Center, The Emergency Management Department, Beijing

[2] School of Remote Sensing and Information Engineering, Wuhan University, Wuhan

来源：

Wuhan Daxue Xuebao (Xinxi Kexue Ban)/Geomatics and Information Science of Wuhan University | 2024年 / 49卷 / 09期

关键词：

BERT; disaster emergency response; earthquake disaster; multi-label classification; pre-trained model; social media; transfer learning;

D O I：

10.13203/j.whugis20220167

中图分类号：

学科分类号：

摘要：

Objectives: With the rapid development of the Internet, social media has become an important information source of emergency events. However, there are a lot of duplication, errors and even malicious contents in social media, which need to be effectively classified to provide more accurate information for di⁃ saster emergency response. Methods: Deep learning has greatly improved the accuracy and efficiency of text classification. This paper takes earthquake disaster as an example, and builds a multi-label classifica⁃ tion model based on bidirectional encoder representation from transformers (BERT) transfer learning. Over 50 000 posts about 5 earthquakes are collected as training samples from SINA Weibo, which is a very popu⁃ lar social media in China. Each sample is manually marked as one or more labels, such as hazards informa⁃ tion, loss information, rescue information, public opinion information and useless information. Results: By fine-tune training, the classification accuracies of the proposed model on training dataset and test dataset reach 97% and 92%, respectively. The area under curve score of each label ranges from 0.952 to 0.998. Conclusions: The results prove that the multi-label classification using BERT transfer learning is of high reliability. The proposed model can be applied to the emergency management services for earthquake events, which is beneficial for the rapid disaster rescue and relief. © 2024 Editorial Department of Geomatics and Information Science of Wuhan University. All rights reserved.

引用

页码：1661 / 1671

页数：10

共 40 条

[21] Yuan Ping, Research on Clustering and Text Classi⁃ fication Based on Support Vector Machine［D］, (2012)
[22] Tongrui Yu, Ran Jin, Xiaozhen Han, Et al., Review of Pre-training Models for Natural Language Pro⁃ cessing［J］, Computer Engineering and Applications, 56, 23, pp. 12-22, (2020)
[23] Nguyen D T, Et al., Applications of Online Deep Learning for Crisis Response Using Social Media Information［EB/OL］
[24] Shuhan Liu, Yandong Wang, Xiaokang Fu, Ex⁃ tracting Rainstorm Disaster Information from Micro⁃ blogs Using Convolutional Neural Network ［J］, Journal of Geo-Information Science, 21, 7, pp. 1009-1017, (2019)
[25] Tengfei Yang, Jibo Xie, Dongchuan Yan, Et al., Ex⁃ tracting Sentiment Information from Social Media Based on Deep Learning and the Research on Disas⁃ ter Reduction［J］, Geography and Geo-Information Science, 36, 2, pp. 62-68, (2020)
[26] Zhoujun Li, Yu Fan, Xianjie Wu, Survey of Natural Language Processing Pre-training Techniques［J］, Computer Science, 47, 3, pp. 162-173, (2020)
[27] Devlin J, Et al., BERT：Pretraining of Deep Bidirectional Transformers for Lan⁃ guage Understanding ［EB/OL］
[28] Peters M E，, Neumann M, Et al., Deep Contextualized Word Representations ［EB/OL］
[29] Im⁃ proving Language Understanding by Generative Pre-training［EB/OL］
[30] Heidari M, Jones J H., Using BERT to Extract Topic-Independent Sentiment Features for Social Media Bot Detection, The 11th IEEE Annual Ubiquitous Computing，Electronics & Mobile Com⁃ munication Conference （UEMCON）, (2020)

← 1 2 3 4 →