Adversarial stylometry
ID: adversarial-stylometry
Adversarial stylometry is a subfield of stylometry, which is the study of linguistic style and the analysis of writing style in texts. Stylometry typically involves identifying and quantifying the distinctive stylistic features of an author’s writing in order to attribute texts to specific authors or to detect plagiarism. In the context of adversarial stylometry, researchers explore techniques that use adversarial machine learning methods to evaluate how robust stylistic features are against manipulations aimed at obfuscating authorship.
New to topics? Read the docs here!