As urban development accelerates and natural disasters occur more frequently, the urgency of developing effective emergency shelter planning strategies intensifies. The shelter location selection method under the traditional multi-criteria decision-making framework suffers from issues such as strong subjectivity and insufficient data support. Artificial intelligence offers a robust data-driven approach for site selection; however, many methods neglect the spatial relationships of site selection targets within geographical space. This paper introduces an emergency shelter site selection model that combines a variational graph autoencoder (VGAE) with a random forest (RF), namely VGAE-RF. In the constructed urban spatial topological graph, based on network geographic information, this model captures both the latent features of geographic unit coupling and integrates explicit and latent features to forecast the likelihood of emergency shelters in the construction area. This study takes Beijing, China, as the experimental area and evaluates the reliability of different model methods using a confusion matrix, Receiver Operating Characteristic (ROC) curve, and Imbalance Index of spatial distribution as evaluation indicators. The experimental results indicate that the proposed VGAE-RF model method, which considers spatial semantic associations, displays the best reliability.

The extraction of disaster geospatial intelligence (DGI) from social media data with spatiotemporal attributes plays a crucial role in real-time disaster monitoring and emergency decision-making. However, conventional machine learning approaches struggle with semantic complexity and limited Chinese disaster corpus. Recent advancements in large language models (LLMs) offer new opportunities to overcome these challenges due to their enhanced semantic comprehension and multi-task learning capabilities. This study investigates the potential application of LLMs in disaster intelligence extraction and proposes an efficient, scalable method for multi-hazard DGI extraction. Building upon a unified ontological framework encompassing core natural disaster elements, this method employs parameter-efficient low-rank adaptation (LoRA) fine-tuning to optimize open-source Chinese LLMs using a meticulously curated instruction-tuning dataset. It achieves simultaneous identification of multi-hazard intelligence cues and extraction of disaster spatial entity attributes from unstructured Chinese social media texts through unified semantic parsing and structured knowledge mapping. Compared to pre-trained models such as BERT and ERNIE, the proposed method was shown to achieve state-of-the-art evaluation results, with the highest recognition accuracy (F1-score: 0.9714) and the best performance in structured information generation (BLEU-4 score: 92.9649). Furthermore, we developed and released DGI-Corpus, a Chinese instruction-tuning dataset covering various disaster types, to support the research and application of LLMs in this field. Lastly, the proposed method was applied to analyze the spatiotemporal evolution patterns of the Zhengzhou “7.20” flood disaster. This study enhances the efficiency of natural disaster monitoring and emergency management, offering technical support for disaster response and mitigation decision-making.

Narrative mapping, an advanced geographic information visualization technology, presents spatial information episodically, enhancing readers’ spatial understanding and event cognition. However, during 3D scene construction, viewpoint selection is heavily reliant on the cartographer’s subjective interpretation of the event. Even with fixed-angle settings, the task of ensuring that selected viewpoints align with the narrative theme remains challenging. To address this, an automated viewpoint selection method constrained by narrative relevance and visual information is proposed. Narrative relevance is determined by calculating spatial distances between each element and the thematic element within the scene. Visual information is quantified by assessing the visual salience of elements as the ratio of their projected area on the view window to their total area. Pearson’s correlation coefficient is used to evaluate the relationship between visual salience and narrative relevance, serving as a constraint to construct a viewpoint fitness function that integrates the visual salience of the convex polyhedron enclosing the scene. The chaotic particle swarm optimization (CPSO) algorithm is utilized to locate the viewpoint position while maximizing the fitness function, identifying a viewpoint meeting narrative and visual salience requirements. Experimental results indicate that, compared to the maximum projected area method and fixed-value method, a higher viewpoint fitness is achieved by this approach. The narrative views generated by this method were positively recognized by approximately two-thirds of invited professionals. This process aligns effectively with narrative visualization needs, enhances 3D narrative map creation efficiency, and offers a robust strategy for viewpoint selection in 3D scene-based narrative mapping.