Data for: Semantics-Enhanced Assisted Evacuation Simulation in Nursing Homes: A Knowledge Graph-Driven Approach
Description
This dataset supports the manuscript “Modeling Assisted Evacuation in Nursing Homes Under Semantic Constraints: A Knowledge Graph-Driven Approach”. It contains the data and materials used to construct and evaluate the proposed knowledge graph-driven assisted evacuation modeling framework. The dataset includes the domain ontology developed in Protégé, the exported nursing home evacuation knowledge graph, simulation parameter settings, dispatch strategy definitions, aggregated experimental results, and source data for the sensitivity analysis figures. Specifically, the files provide the semantic structure of residents, caregivers, assistive devices, spatial units, and their relations, as well as the numerical results used to generate the main performance comparison and sensitivity analyses reported in the manuscript. Raw facility information used to construct the evacuation scenario is not included due to confidentiality restrictions. The shared files provide the processed ontology, graph structure, model parameters, and aggregated simulation results necessary to understand and verify the findings of the study.
Files
Steps to reproduce
1. Open the ontology file NursingHomeEvacuation.owl in Protégé to inspect the domain classes, attributes, and semantic relations used for nursing-home evacuation modeling. 2. Use the exported knowledge graph file knowledge_graph/graph_backup.json, knowledge_graph/edges.csv, knowledge_graph/nodes.csv, knowledge_graph/state.csv to examine the instantiated semantic entities and relations, including residents, caregivers, assistive devices, spatial units, and assistance dependencies. 3. Refer to model_parameters.xlsx for the simulation settings, including time step, corridor capacity, task-start budget, reaction-delay distribution, population configuration, and repeated-run settings. 4. Refer to strategy_settings.xlsx for the definitions of the four dispatch strategies: nearest dispatch, stretcher-priority, floor-priority, and flow-control. 5. Use main_results_Table6.xlsx to verify the aggregated experimental results reported in Table 6 of the manuscript. 6. Use Fig4_source_data.xlsx and Fig5_source_data.xlsx to reproduce the sensitivity analysis figures for task-start budget and corridor capacity, respectively. Raw facility information used to construct the evacuation scenario is not included due to confidentiality restrictions. The shared files provide the processed ontology, graph structure, parameter settings, and aggregated simulation results needed to understand and verify the findings.
Institutions
- University of Science and Technology BeijingBeijing, Beijing
- Shanghai Jiao Tong UniversityShanghai, Shanghai