journal article·2026-02-05Open access

HiDEF: A Hierarchical Disaster Information Extraction Framework Based on Adversarial Augmentation and Dynamic Prompting

Xiaodong Wang, Tengfei Yang, Xiaohan Yang

Applied Sciences· Volume 16· Issue 3

Abstract

In disaster emergency response, spatial location information embedded within social media texts holds substantial value for the rapid localization of affected areas and the implementation of precise rescue operations. Existing research predominantly employs natural language processing and deep learning technologies for geographic information extraction; however, two critical limitations persist: first, insufficient integration of textual semantic features for disaster relevance determination, resulting in inadequate correlation between extracted results and actual disaster locations; second, absence of mechanisms for identifying affected sites in multi-location contexts, thereby compromising decision support efficacy. Addressing these challenges, this study proposes a hierarchical disaster location information extraction framework that integrates semantic understanding. The framework operates through a three-tier hierarchy: data-level adversarial augmentation, semantic-level dynamic parsing, and parameter-level scale optimization. It achieves three core functionalities: (1) precise determination of disaster relevance for geographic location information; (2) identification of affected areas in multi-location contexts; (3) establishment of a logarithmic scaling relationship between LLM parameter scale and optimal prompt sample size.
Source: crossref · Quality: baseline

Research context

OpenAlex topics: Public Relations and Crisis Communication · Geographic Information Systems Studies · Data-Driven Disease Surveillance

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

School of Mathematics and Statistics, Henan University of Science and Technology, Luoyang 471023, China

Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

Institute of Information Technology, Information Engineering University, Zhengzhou 450002, China

References (27)

Show source-provided references
1. Gao, S., Yang, T., Xu, Y., Mou, N., Wang, X., and Huang, H. (2025). Enhancing Disaster Situation Awareness Through Multimodal Social Media Data: Evidence from Typhoon Haikui. Appl. Sci., 15.10.3390/app15010465 ↗
2. Processing Social Media Messages in Mass Emergency: A Survey10.1145/2771588 ↗
3. Understanding social media data for disaster management10.1007/s11069-015-1918-0 ↗
4. Rapid assessment of disaster damage using social media activity10.1126/sciadv.1500779 ↗
5. Combining machine-learning topic models and spatiotemporal analysis of social media data for disaster footprint and damage assessment10.1080/15230406.2017.1356242 ↗
6. Rapid flood inundation mapping using social media, remote sensing and topographic data10.1007/s11069-017-2755-0 ↗
7. Sloan, L., Morgan, J., Burnap, P., and Williams, M. (2015). Who Tweets? Deriving the Demographic Characteristics of Age, Occupation and Social Class from Twitter User Meta-Data. PLoS ONE, 10.10.1371/journal.pone.0115545 ↗
8. Real-Time Crisis Mapping of Natural Disasters Using Social Media10.1109/MIS.2013.126 ↗
9. Malik, M.M., Lamba, H., Nakos, C., and Pfeffer, J. (2015, January 26–29). Population Bias in Geotagged Tweets. Proceedings of the International AAAI Conference on Web and Social Media, Oxford, UK.
10. Spatial biases in crowdsourced data: Social media content attention concentrates on populous areas in disasters10.1016/j.compenvurbsys.2020.101514 ↗
11. Dhakal, A. (2025). Mitigation of Spatial Bias in Social Media Data for Disaster Relief. [Master’s Thesis, University of Wisconsin-Milwaukee].
12. Opportunities and risks of disaster data from social media: A systematic review of incident information10.5194/nhess-21-1431-2021 ↗
13. Extraction and analysis of natural disaster-related VGI from social media: Review, opportunities and challenges10.1080/13658816.2022.2048835 ↗
14. Research on Named Entity Recognition Methods for Urban Underground Space Disasters Based on Text Information Extraction10.5194/isprs-archives-XLVIII-1-W2-2023-547-2023 ↗
15. Deep Learning-Based Methods for Natural Hazard Named Entity Recognition10.1038/s41598-022-08667-2 ↗
16. Fine-tuned BERT-BiLSTM-CRF approach for named entity recognition in geological disaster texts10.1007/s12145-025-01870-5 ↗
17. Leveraging Social Media for Managing Natural Hazard Disasters: A Critical Review of Data Collection Strategies and Actionable Insights
18. Hameed, S.N., Ranathunga, S., Prasanna, R., Stock, K., and Jones, C.B. (2025). Extracting Disaster Impacts and Impact Related Locations in Social Media Posts Using Large Language Models. arXiv.
19. Multi-Class Multi-Label Classification of Social Media Texts for Typhoon Damage Assessment: A Two-Stage Model Fully Integrating the Outputs of the Hidden Layers of BERT10.1080/17538947.2024.2348668 ↗
20. Morris, J.X., Lifland, E., Yoo, J.Y., Grigsby, J., Jin, D., and Qi, Y. (2020, January 16–20). TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, Online.10.18653/v1/2020.emnlp-demos.16 ↗
21. Guo, D., Yang, D., Zhang, H., Song, J., Zhang, R., Xu, R., Zhu, Q., Ma, S., Wang, P., and Bi, X. (2025). Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv.
22. Qin, W., Suo, L., Li, L., and Yang, F. (2025). Advancing Software Vulnerability Detection with Reasoning LLMs: DeepSeek-R1’s Performance and Insights. Appl. Sci., 15.10.3390/app15126651 ↗
23. Goldstein, J., and Carbonell, J.G. (1998). Summarization:(1) using MMR for diversity-based reranking and (2) evaluating summaries. TIPSTER Text Program Phase III: Proceedings of a Workshop Held at Baltimore, Maryland, 13–15 October 1998, Association for Computational Linguistics.
24. DeepSeek-AI (2025, May 04). Deepseek-Ai/DeepSeek-R1 GitHub Repository. GitHub 2025. Available online: https://github.com/deepseek-ai/DeepSeek-R1.
25. Nussbaum, Z., Morris, J.X., Duderstadt, B., and Mulyar, A. (2024). Nomic embed: Training a reproducible long context text embedder. arXiv.
26. Ye, X., Iyer, S., Celikyilmaz, A., Stoyanov, V., Durrett, G., and Pasunuru, R. (2023, January 9–14). Complementary explanations for effective in-context learning. Proceedings of the Findings of the Association for Computational Linguistics: ACL 2023, Toronto, ON, Canada.10.18653/v1/2023.findings-acl.273 ↗
27. SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python10.1038/s41592-019-0686-2 ↗