Effective writing style transfer via combinatorial paraphrasing

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

116 Downloads (Pure)

Abstract

Stylometry can be used to profile or deanonymize authors against their will based on writing style. Style transfer provides a defence. Current techniques typically use either encoder-decoder architectures or rule-based algorithms. Crucially, style transfer must reliably retain original semantic content to be actually deployable. We conduct a multifaceted evaluation of three state-of-the-art encoder-decoder style transfer techniques, and show that all fail at semantic retainment. In particular, they do not produce appropriate paraphrases, but only retain original content in the trivial case of exactly reproducing the text. To mitigate this problem we propose ParChoice: a technique based on the combinatorial application of multiple paraphrasing algorithms. ParChoice strongly outperforms the encoder-decoder baselines in semantic retainment. Additionally, compared to baselines that achieve non-negligible semantic retainment, ParChoice has superior style transfer performance. We also apply ParChoice to multi-author style imitation (not considered by prior work), where we achieve up to 75% imitation success among five authors. Furthermore, when compared to two state-of-the-art rule-based style transfer techniques, ParChoice has markedly better semantic retainment. Combining ParChoice with the best performing rule-
based baseline (Mutant-X [34]) also reaches the highest style transfer success on the Brennan-Greenstadt and Extended-Brennan-Greenstadt corpora, with much less impact on original meaning than when using the rule-based baseline techniques alone. Finally, we highlight a critical problem that afflicts all current style transfer techniques: the adversary can use the same technique for thwarting style transfer via adversarial training. We show that adding randomness to style transfer helps to mitigate the effectiveness of adversarial training.
Original languageEnglish
Title of host publicationProceedings on Privacy Enhancing Technologies
Pages175-195
Number of pages20
DOIs
Publication statusPublished - 17 Aug 2020
MoE publication typeA4 Article in a conference publication
EventPrivacy Enhancing Technologies Symposium - Montreal, Canada
Duration: 15 Jul 202019 Jul 2020
Conference number: 20

Publication series

NameProceedings on Privacy Enhancing Technologies
PublisherDe Gruyter
Number4
Volume2020
ISSN (Electronic)2299-0984

Conference

ConferencePrivacy Enhancing Technologies Symposium
Abbreviated titlePETS
CountryCanada
CityMontreal
Period15/07/202019/07/2020

Keywords

  • style transfer
  • style imitation
  • stylometry
  • adversarial stylometry
  • author profiling
  • profiling
  • deanonymization
  • model evasion

Fingerprint Dive into the research topics of 'Effective writing style transfer via combinatorial paraphrasing'. Together they form a unique fingerprint.

Cite this