Fine-grained flood disaster information extraction incorporating multiple semantic features

Flood disasters rank as the most prevalent natural calamities of the twenty-first century, incurring extensive human and economic losses globally. As a crucial source for disaster monitoring, social media data exhibits high variability and ambiguity, with current research lacking targeted multidimen...

Full description

Saved in:
Bibliographic Details
Main Authors: Shunli Wang, Rui Li, Huayi Wu, Jiang Li, Yun Shen
Format: Article
Language:English
Published: Taylor & Francis Group 2025-08-01
Series:International Journal of Digital Earth
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/17538947.2024.2448221
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849224323202547712
author Shunli Wang
Rui Li
Huayi Wu
Jiang Li
Yun Shen
author_facet Shunli Wang
Rui Li
Huayi Wu
Jiang Li
Yun Shen
author_sort Shunli Wang
collection DOAJ
description Flood disasters rank as the most prevalent natural calamities of the twenty-first century, incurring extensive human and economic losses globally. As a crucial source for disaster monitoring, social media data exhibits high variability and ambiguity, with current research lacking targeted multidimensional semantic analysis, resulting in coarse granularity and limited accuracy. To address this problem, this study proposes a framework and method synthesizing multiple semantic features to extract fine-grained disaster information. Static embeddings representing stable semantics and dynamic embeddings representing changing semantics are fused to extract toponyms, with the depth-first search used to generate addresses through the toponym tree. Guiding prompts incorporating domain-specific knowledge of disasters are designed for the large language model, with an iterative feedback process refining location-based disaster information. Finally, the reliability of social media-sourced information is assessed by comparing extracted flooded locations with actual monitoring data. The case study on the Zhengzhou ‘7·20’ flood event demonstrates the effectiveness of our semantic fusion approach, achieving an F1 score of 0.9384 for address extraction, with accuracies of 0.8485 and 0.8788 for waterlogging depth and trapped individuals, respectively. This research offers a practical framework for nuanced perception and timely rescue operations in urban disaster management.
format Article
id doaj-art-caea5775ea6c408a9eaef8a60a46612d
institution Kabale University
issn 1753-8947
1753-8955
language English
publishDate 2025-08-01
publisher Taylor & Francis Group
record_format Article
series International Journal of Digital Earth
spelling doaj-art-caea5775ea6c408a9eaef8a60a46612d2025-08-25T11:31:49ZengTaylor & Francis GroupInternational Journal of Digital Earth1753-89471753-89552025-08-0118110.1080/17538947.2024.2448221Fine-grained flood disaster information extraction incorporating multiple semantic featuresShunli Wang0Rui Li1Huayi Wu2Jiang Li3Yun Shen4State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, People’s Republic of ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, People’s Republic of ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, People’s Republic of ChinaInformation Center of Department of Natural Resources of Hubei Province, Wuhan, People’s Republic of ChinaInformation Center of Department of Natural Resources of Hubei Province, Wuhan, People’s Republic of ChinaFlood disasters rank as the most prevalent natural calamities of the twenty-first century, incurring extensive human and economic losses globally. As a crucial source for disaster monitoring, social media data exhibits high variability and ambiguity, with current research lacking targeted multidimensional semantic analysis, resulting in coarse granularity and limited accuracy. To address this problem, this study proposes a framework and method synthesizing multiple semantic features to extract fine-grained disaster information. Static embeddings representing stable semantics and dynamic embeddings representing changing semantics are fused to extract toponyms, with the depth-first search used to generate addresses through the toponym tree. Guiding prompts incorporating domain-specific knowledge of disasters are designed for the large language model, with an iterative feedback process refining location-based disaster information. Finally, the reliability of social media-sourced information is assessed by comparing extracted flooded locations with actual monitoring data. The case study on the Zhengzhou ‘7·20’ flood event demonstrates the effectiveness of our semantic fusion approach, achieving an F1 score of 0.9384 for address extraction, with accuracies of 0.8485 and 0.8788 for waterlogging depth and trapped individuals, respectively. This research offers a practical framework for nuanced perception and timely rescue operations in urban disaster management.https://www.tandfonline.com/doi/10.1080/17538947.2024.2448221Flood disasterssocial mediafusion embeddingknowledge-guidedlarge language model fine-tuningdisaster information extraction
spellingShingle Shunli Wang
Rui Li
Huayi Wu
Jiang Li
Yun Shen
Fine-grained flood disaster information extraction incorporating multiple semantic features
International Journal of Digital Earth
Flood disasters
social media
fusion embedding
knowledge-guided
large language model fine-tuning
disaster information extraction
title Fine-grained flood disaster information extraction incorporating multiple semantic features
title_full Fine-grained flood disaster information extraction incorporating multiple semantic features
title_fullStr Fine-grained flood disaster information extraction incorporating multiple semantic features
title_full_unstemmed Fine-grained flood disaster information extraction incorporating multiple semantic features
title_short Fine-grained flood disaster information extraction incorporating multiple semantic features
title_sort fine grained flood disaster information extraction incorporating multiple semantic features
topic Flood disasters
social media
fusion embedding
knowledge-guided
large language model fine-tuning
disaster information extraction
url https://www.tandfonline.com/doi/10.1080/17538947.2024.2448221
work_keys_str_mv AT shunliwang finegrainedflooddisasterinformationextractionincorporatingmultiplesemanticfeatures
AT ruili finegrainedflooddisasterinformationextractionincorporatingmultiplesemanticfeatures
AT huayiwu finegrainedflooddisasterinformationextractionincorporatingmultiplesemanticfeatures
AT jiangli finegrainedflooddisasterinformationextractionincorporatingmultiplesemanticfeatures
AT yunshen finegrainedflooddisasterinformationextractionincorporatingmultiplesemanticfeatures