Geological disaster causes analysis based on knowledge graph link prediction

Xinya Lei (), Changle Li (), Yuewei Wang () and Weijing Song ()
Additional contact information
Xinya Lei: China University of Geosciences
Changle Li: China University of Geosciences
Yuewei Wang: China University of Geosciences
Weijing Song: China University of Geosciences

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards, 2025, vol. 121, issue 18, No 28, 21483-21504

Abstract: Abstract Geological disasters have complicated natural and man-made causes. Identifying the event’s causes aids disaster management. The existing model mainly uses geospatial data to analyze the spatial-temporal correlation among various factors contributing to geological disasters in the region, ignoring the valuable causal information in disaster-related texts, making it impossible to identify the causes of each event. To address this challenge, we have constructed a geological disaster knowledge graph (GDKG) that integrates geospatial and textual data. This GDKG comprises geological disaster event entities, disaster-inducing factor entities, causal similarity relations among event entities, and causal relationships between events and their inducing factors. We also present Attention-ComplEx, an inductive link prediction model. This model overcomes the limitation of conventional inductive link prediction models, which struggle to predict links rely solely on entity attributes without information about the surrounding context. Our model uses attention mechanisms to capture the interactions between disaster entities’ influencing factors and their importance in characterizing their causal features to generate embeddings for disaster event entities and the ComplEx model to learn GDKG entity and relationship correlations. Predicting potential links between disaster event entities and disaster-inducing factors allows us to effectively identify disaster event causes. Utilizing multi-source geospatial and textual data spanning from 2011 to 2022 in Hunan, Hubei, and Jiangxi provinces, we have constructed a knowledge graph and a link prediction dataset. Experimental results demonstrate that the Attention-ComplEx model outperforms baseline methods in link prediction with the highest MRR. The Attention-ComplEx model identifies disaster causes better than SVM and XGBOOST model.

Keywords: Geological disasters; Natural and man-made factors; Geological disaster causes identification; Knowledge graph; Link prediction (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11069-025-07642-0 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:nathaz:v:121:y:2025:i:18:d:10.1007_s11069-025-07642-0

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11069

DOI: 10.1007/s11069-025-07642-0

Access Statistics for this article

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards is currently edited by Thomas Glade, Tad S. Murty and Vladimír Schenk

More articles in Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards from Springer, International Society for the Prevention and Mitigation of Natural Hazards
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().