Driving Behavior Dataset for Driver Identification and Anti-theft Analysis

This dataset contains real-world driving data collected from ten drivers operating a production vehicle along a fixed route between Korea University and Sangam World Cup Stadium. The data covers approximately 23 hours of driving over a round-trip distance of 46 km. Driving logs were recorded once per second and include 51 in-vehicle sensor measurements obtained through OBD-II. The dataset provides labeled driving traces for drivers A to J and supports research on driver identification, driving-style modeling, behavioral biometrics, and vehicle anti-theft analysis. It forms the basis of studies using mechanical, fuel-related, and behavioral features extracted from in-vehicle sensor data for machine-learning-based driver classification.

本数据集收录了10名驾驶员驾驶量产车辆，沿高丽大学（Korea University）至上岩世界杯体育场（Sangam World Cup Stadium）的固定路线采集的真实驾驶数据。该数据集涵盖约23小时的驾驶记录，往返行程总距离达46公里。驾驶日志以每秒1次的频率进行记录，包含通过车载诊断系统第二代（OBD-II）获取的51项车载传感器测量数据。本数据集为A至J号驾驶员提供了带标注的驾驶轨迹，可支撑驾驶员身份识别、驾驶风格建模、行为生物识别以及车辆防盗分析等相关研究。其为利用车载传感器数据提取的机械特征、燃油相关特征与行为特征，开展基于机器学习的驾驶员分类研究提供了核心实验基础。