Jokestega: automatic joke generation-based steganography methodology

摘要

This paper presents a novel steganography methodology, namely Automatic Joke Generation Based Steganography Methodology (Jokestega), that pursues textual jokes in order to hide messages. Basically, Jokestega methodology takes advantage of recent advances in Automatic Jokes Generation (AJG) techniques to automate the generation of textual steganographic cover. In a corpus of jokes, one may judge a number of documents to be the same joke although letters, locations, and other details are different. Generally, joke and puns could be retold with totally different vocabulary, while still retaining their identities. Therefore, Jokestega pursues the common variations among jokes to conceal data. Furthermore, when someone is joking, anything may be said which legitimises the use of joke-based steganography. This makes employing textual jokes very attractive as steganographic carrier for camouflaging data. It is worth noting that Jokestega follows Nostega paradigm, which implies that joke-cover is noiseless. The validation results demonstrate the effectiveness of Jokestega.