DT-CREATE Replication Package: A Service Suite for Self-Adaptive Digital Twins Supporting Predictive Maintenance Decisions in Industry 5.0

Replication package for the paper "DT-CREATE: A Service Suite for Self-Adaptive Digital Twins Supporting Predictive Maintenance Decisions in Industry 5.0", focused on decision support for predictive maintenance. Contains five datasets (textile industry, heat-treatment furnace, NASA C-MAPSS FD001, and de-identified industry machine logs), Python scripts reproducing all classification experiments (ADAPTflow/PyCaret, single-algorithm baselines, framework-level baselines, and ontology/SWRL ablation analysis), two Jupyter notebooks from the exploratory workflow, and JSON result files with per-seed metrics reported in the paper. The raw industry export (~613 MB, confidential) is not redistributed; all included files are sufficient to reproduce the reported results.