Distinguishing a Human or Machine Cyberattacker

A Linguistic Analysis Metric in Detecting Ransomware Cyber-attacks

“…www.ijacsa.thesai.org The average MLD values, emerging from the translation into one of the available languages and then back to English for each of the test sentences, are provided in Table IV. The sources for the quotations can be found at [8]. This examination is intended to ascertain the cyberattack's original source language.…”

“…In previous studies [7,8], we have analyzed ransomware external messaging via a mechanism that has involved six extensively used languages (French, Spanish, German, Russian, Chinese, and Hindi) and several round-trip translation (RTT) operations from target language into English using the Google Translate (GT) functionality. Our analysis was then conducted on a number of random quotations from popular culture and English literature and that finally allowed us to devise a procedure which could establish whether a perceived attack was initiated by a human writer with some knowledge of English, or alternatively, by a machine translation.…”

“…Comparisons are made in this study among the six chosen Eastern European region nations, but the linguistic differences among these nations and their natural languages are somewhat obscured when analyzed in the context of these languages compared to other world languages where the linguistic structures are so different. For example, in an earlier paper [8] we considered more closely related European-based languages compared to Hindi, Russian and Chinese. In order to draw greater distinctions between the set of languages in this study, we used the same test bed (T1-T20) of well-known English language text, and sought to find a subset of the test items that would provide a greater discrimination between the Eastern European languages in the study.…”

“…Cyberattacks may be motivated by ideological, financial, or personal reasons and are directed at governments and institutions, businesses and private individuals engendering geopolitical, security, reputational and privacy concerns. While there is a wide literature on the typology, counter measures, policies and security information sharing across state and private sectors [1][2][3][4][5][6][7][8], for our practical purposes, we shall briefly refer in this section to ransomware and ways to address cybersecurity by means of linguistic approaches and instruments.…”

“…With ransomware, linguistic approaches have been either in the form of actual text-analyses (see [8]) or of ransomware detection devices and apps based on linguistic parsing [17]. Two types of linguistic analysis can be distinguished: one that examines the way the source code was written and another that examines the text that was used.…”