Deep excavation construction risk assessment dataset for a metro station in Chongqing

This deposit provides a reusable deep excavation construction risk assessment package for a metro station case in Chongqing, China. It supports method benchmarking, teaching, sensitivity analysis, and adaptation to additional excavation projects that use the same indicator schema. Package contents- data/: tabular files 00–10 (CSV/JSON), including project metadata, a 27-indicator dictionary (G1–G4, T1–T5, E1–E5, M1–M4, O1–O5, H1–H4), five-level grade boundaries, standard normal cloud parameters (Ex, En, He), raw scores from three expert groups, aggregated case indicator values, combined subjective–objective weights (alpha = 0.6), expert consistency metrics (Fleiss’ kappa and ICC), narrative grade criteria, and reference association/evaluation outputs- code/: batch-runnable MATLAB scripts (R2018b or later) for dataset loading, Monte Carlo extensible cloud evaluation (default T = 1000, seed 2024), hyper-entropy sensitivity analysis, and automated validation- docs/: data dictionary- results/: example validation and evaluation outputs from the reference pipeline Case summaryIsland-type metro station excavation: length 139.6 m, average depth 32 m, width 22.8 m; supported by cast-in-place bored piles and steel internal supports. Twenty-two indicators are numeric and five are intervals (E4, M4, O4, H3, H4). Licence- Data files: Creative Commons Attribution 4.0 International (CC BY 4.0)- Code: MIT License Please cite this dataset DOI when reusing the materials.