Defending Against Authorship Identification Attacks
Not on the map. Why?
Abstract (arXiv)
Authorship identification has proven unsettlingly effective in inferring the identity of the author of an unsigned document, even when sensitive personal information has been carefully omitted. In the digital era, individuals leave a lasting digital footprint through their written content, whether it is posted on social media, stored on their employer's computers, or located elsewhere. When individuals need to communicate publicly yet wish to remain anonymous, there is little available to protect them from unwanted authorship identification. This unprecedented threat to privacy is evident in scenarios such as whistle-blowing. Proposed defenses against authorship identification attacks primarily aim to obfuscate one's writing style, thereby making it unlinkable to their pre-existing writing, while concurrently preserving the original meaning and grammatical integrity. The presented work offers a comprehensive review of the advancements in this research area spanning over the past two decades and beyond. It emphasizes the methodological frameworks of modification and generation-based strategies devised to evade authorship identification attacks, highlighting joint efforts from the differential privacy community. Limitations of current research are discussed, with a spotlight on open challenges and potential research avenues.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 34 of its 128 references that are in the map. First posted on arXiv 2 Oct 2023.
Builds on
34 of 128 references.
References · 34
- 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
- 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
- 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
- 2016Deep Learning with Differential Privacycited by 2,682
- 2015A large annotated corpus for learning natural language inferencecited by 1,837
- 2019What Does BERT Look At? An Analysis of BERT's Attentioncited by 994
- 2018Neural Network Acceptability Judgmentscited by 952
- 2020Beyond English-Centric Multilingual Machine Translationcited by 559
- 2017Wild Patterns: Ten Years After the Rise of Adversarial Machine Learningcited by 411
- 2020Deep Learning for Text Style Transfer: A Surveycited by 161
- 2016Counter-fitting Word Vectors to Linguistic Constraintscited by 138
- 2020CIAGAN: Conditional Identity Anonymization Generative Adversarial Networkscited by 88
22 more
- 2023White-Box Transformers via Sparse Rate Reductioncited by 77
- 2022The Limits of Word Level Differential Privacycited by 63
- 2017A4NT: Author Attribute Anonymity by Adversarial Training of Neural Machine Translationcited by 41
- 2019Speaker Anonymization Using X-vector and Neural Waveform Modelscited by 41
- 2019ER-AE: Differentially Private Text Generation for Authorship Anonymizationcited by 35
- 2023DP-BART for Privatized Text Rewriting under Local Differential Privacycited by 27
- 2017Evaluating prose style transfer with the Biblecited by 23
- 2021Privacy-Preserving Graph Convolutional Networks for Text Classificationcited by 19
- 2021Idiosyncratic but not Arbitrary: Learning Idiolects in Online Registers Reveals Distinctive yet Consistent Individual Stylescited by 18
- 2022Text Revision by On-the-Fly Representation Optimizationcited by 17
- 2017The Case for Being Average: A Mediocrity Approach to Style Masking and Author Obfuscationcited by 15
- 2021Counterfactual Evaluation for Explainable AIcited by 15
- 2022Are You Robert or RoBERTa? Deceiving Online Authorship Attribution Models Using Neural Text Generatorscited by 14
- 2018Style Obfuscation by Invariancecited by 10
- 2021DeepStyle: User Style Embedding for Authorship Attribution of Short Textscited by 10
- 2014Fighting Authorship Linkability with Crowdsourcingcited by 8
- 2021Avengers Ensemble! Improving Transferability of Authorship Obfuscationcited by 7
- 2018Author Obfuscation Using Generalised Differential Privacycited by 4
- 2021TranslateLocally: Blazing-fast translation running on the local CPUcited by 3
- 2022A Girl Has A Name, And It's ... Adversarial Authorship Attribution for Deobfuscationcited by 3
- 2015From Closed-world Enforcement to Open-world Assessment of Privacycited by 2
- 2022Reproduction and Replication of an Adversarial Stylometry Experimentcited by 2
Built on by
Citing papers per year: 2024: 1.
1 paper.