知识卡片

普通读书笔记卡 · 8836 · 生成模型:GPT-6

匿名化可能被外部信息关联打破

理解等级:中级

去掉直接标识符后,数据仍可能通过与其它来源匹配而重新识别人。隐私不能只以单一数据集是否含姓名判断,还要考虑发布内容和可关联信息。原书历史重识别故事用于说明机制,不据此补写当前个案法律或事实结论。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      seemingly cleaned data, when combined with another apparently innocuous dataset, can reveal information that was desired to be private. For example, in 1997, the medical records of the governor of Massachussetts were identified by matching anonymized medical data with publicly available voter registration records [254, 111]. The combination of the two datasets allowed seemingly de-identified private data to be re-identified.

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第04章《计算机视觉与社会》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c004.xhtml) - 结论依据:摘录中关于匿名化可能被外部信息关联打破的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:seemingly cleaned data, when combined with another apparently innocuous dataset, can reveal information that was desired to be private. For example, in 1997, the medical records of the governor of Massachussetts were identified by matching anonymized medical data with publicly available voter registration records [254, 111]. The combination of the two datasets allowed seemingly de-identified private data to be re-identified.