English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent advances in natural language processing have enabled powerfulprivacy-invasive authorship attribution. To counter authorship attribution,researchers have proposed a variety of rule-based and learning-based textobfuscation approaches. However, existing authorship obfuscation approaches donot consider the adversarial threat model. Specifically, they are not evaluatedagainst adversarially trained authorship attributors that are aware ofpotential obfuscation. To fill this gap, we investigate the problem ofadversarial authorship attribution for deobfuscation. We show thatadversarially trained authorship attributors are able to degrade theeffectiveness of existing obfuscators from 20-30% to 5-10%. We also evaluatethe effectiveness of adversarial training when the attributor makes incorrectassumptions about whether and which obfuscator was used. While there is a aclear degradation in attribution accuracy, it is noteworthy that thisdegradation is still at or above the attribution accuracy of the attributorthat is not adversarially trained at all. Our results underline the need forstronger obfuscation approaches that are resistant to deobfuscation

A Girl Has A Name, And It's ... Adversarial Authorship Attribution for Deobfuscation