Authcode - Dataset

AuthCODE is a multi-device continuous authentication architecture that guarantees the privacy of users' sensitive data while provides good authentication performance thanks to an hybrid approach that combines the Mobile Edge Computing (MEC) and Cloud Computing paradigms. A Smart Office scenario where five users interact with their personal computers and smartphones have been used to generate the following five datasets and measure the AuthCODE performance..Dataset 1. Single-device behaviour profile obtained from the personal computer, comprising aggregated data about keyboard and mouse activity, as well as application usage statistics.Dataset 2. Single-device behaviour profile obtained from the mobile device, with application usage statistics.Dataset 3. Single-device behaviour profile with features computed from the sensors of the mobile device.Dataset 4. Multi-device behaviour profile combining the most relevant features of the mobile device and personal computer.Dataset 5. Multi-device behaviour profile generated from the active/inactive intervals of both devices. There are multiple sub-datasets regarding the window size utilised, from 1 hour to 24 hours.

AuthCODE是一种多设备连续身份认证架构，其通过融合移动边缘计算（Mobile Edge Computing, MEC）与云计算范式的混合方案，在保障用户敏感数据隐私的同时，实现了优异的认证性能。
研究采用了包含5名用户与个人电脑、智能手机交互的智能办公场景，以此生成下述5个数据集并对AuthCODE的认证性能进行评测。
数据集1：源自个人电脑的单设备行为画像，包含键盘与鼠标活动的聚合数据，以及应用使用统计信息。
数据集2：源自移动设备的单设备行为画像，仅包含应用使用统计数据。
数据集3：基于移动设备传感器计算得到特征的单设备行为画像。
数据集4：融合移动设备与个人电脑核心特征的多设备行为画像。
数据集5：基于两款设备的活跃/非活跃区间生成的多设备行为画像。此外，该数据集还包含多组基于不同窗口大小（1小时至24小时）的子数据集。