Synthetic IIoT ABAC Policy Dataset: 100K Rules with Zipf-Distributed Topic Access

Synthetic dataset of 100,000 attribute-based access control (ABAC) rules for Industrial IoT (IIoT) authorization benchmarking. Generated with RANDOM_SEED=42 for full reproducibility. The dataset models a K=4 dimensional ABAC policy space: role (5 values), MQTT topic (63,306 unique topics), zone (5 manufacturing zones), and threat level (3 security levels). Topic access frequencies follow a Zipf distribution with exponent s=1.07 (R^2=0.94). Includes generator script, Zipf calibration script, 1,000 conflict rulepairs, and fitted distribution parameters.