BriefGPT.xyz
Aug, 2020
去标识化文本转化的隐私保证
Privacy Guarantees for De-identifying Text Transformations
HTML
PDF
David Ifeoluwa Adelani, Ali Davody, Thomas Kleinbauer, Dietrich Klakow
TL;DR
本文基于差分隐私原理给出了关于基于文本转换的去识别化方法的形式化隐私保证,并探究了不同的掩盖策略对与自然语言处理任务的影响。作者发现,只有通过深度学习模型进行逐字替换的方法是在多个任务中具有鲁棒性的。
Abstract
machine learning
approaches to
natural language processing
tasks benefit from a comprehensive collection of real-life user data. At the same time, there is a clear need for protecting the
→