Authorship attribution

Public presentation by Inari Listenmaa
on Tue. 02 May 2017 at 15:30-17:30 in room EDIT 3364
The way you write text—essays, emails, tweets, even source code—reveals a lot about you, even if you have taken other means to ensure anonymity or pseudonymity. In this talk, we introduce the problem of authorship attribution: how does it work, what can be identified, and if, or how, it can be circumvented. 
View PDF

Introductory papers
  • Brennan et al. Adversarial Stylometry: Circumventing Authorship Recognition to Preserve Privacy and Anonymity (TISSEC 2012)
  • Advanced papers
  • Caliskan et al. De-anonymizing programmers via code stylometry (USENIX 2015)
  • Overdorf et al. Blogs, Twitter Feeds, and Reddit Comments: Cross-domain Authorship Attribution (PoPETS 2016)
  • Fork me on GitHub