Text Analysis in Adversarial Settings: Does Deception Leave a Stylistic Trace?

Abstract

Textual deception constitutes a major problem for online security. Many studies have argued that deceptiveness leaves traces in writing style, which could be detected using text classification techniques. By conducting an extensive literature review of existing empirical work, we demonstrate that while certain linguistic features have been indicative of deception in certain corpora, they fail to generalize across divergent semantic domains. We suggest that deceptiveness as such leaves no content-invariant stylistic trace, and textual similarity measures provide superior means of classifying texts as potentially deceptive. Additionally, we discuss forms of deception beyond semantic content, focusing on hiding author identity by writing style obfuscation. Surveying the literature on both author identification and obfuscation techniques, we conclude that current style transformation methods fail to achieve reliable obfuscation while simultaneously ensuring semantic faithfulness to the original text. We propose that future work in style transformation should pay particular a ention to disallowing semantically drastic changes.

Assessing the Applicability of Authorship Verification Methods

“…In recent years, the new research field author obfuscation (AO) evolved, which concerns itself with the task to fool AA or AV methods in a way that the true author cannot be correctly recognized anymore. To achieve this, AO approaches which, according to Gröndahl and Asokan [7] can be divided into manual, computer-assisted and automatic types, perform a variety of modifications on the texts. These include simple synonym replacements, rule-based substitutions or word order permutations.…”

“…As a first corpus, we compiled C DBLP that represents a collection of 80 excerpts from scientific works including papers, dissertations, book chapters and technical reports, which we have chosen from the well-known Digital Bibliography & Library Project (DBLP) platform 7 . Overall, the documents 8 were written by 40 researchers, where for each author A, there are exactly two documents.…”