Linguistic Steganography using Automatically Acquired Paraphrases
- When?
- Wednesday 25 November 2009, 14:00 to 15:00
- Where?
- 39 BB 02
- Open to:
- Staff, Students, Public
- Speaker:
- Stephen Clark
- Admission price:
- Free
Linguistic Steganography aims to provide techniques for hiding information in natural language texts, through manipulating properties of the text, for example by replacing some words with their synonyms. Unlike image-based steganography, linguistic steganography is in its infancy with little existing work.
In this talk I will motivate the problem, in particular as an interesting application for consideration by both natural language processing and security researchers. Linguistic steganography is a difficult problem because any change to the cover text must retain the meaning and style of the original, in order to prevent detection by an adversary. Our method embeds information in the cover text by replacing phrases in the text with appropriate paraphrases. For part of the talk I will describe how suitable paraphrases can be found automatically by exploiting parallel texts: sequences of sentences with corresponding translations in another language. This automatic method does not always provide accurate paraphrases; for the remainder of the talk I will describe a simple method for determining whether a phrase can be safely substituted with a particular paraphrase given a particular context, together with an empirical evaluation, and I will briefly describe a possible embedding method. Joint work with Ching-Yun (Frannie) Chang (Oxford University Computing Laboratory).
Brief Biography: Stephen Clark is a Senior Lecturer at the University of Cambridge Computer Laboratory, and a member of the Natural Language and Information Processing Research Group. From 2004 to 2008 he was a University Lecturer at the Oxford University Computing Laboratory, and a Tutorial Fellow of Keble College. Before that he spent four years as a postdoctoral researcher at the University of Edinburgh's School of Informatics, working with Mark Steedman. He has a PhD in Artificial Intelligence from the University of Sussex and a first degree in Philosophy from the University of Cambridge. In 2010 Clark is a Programme Co-Chair for the 48th Annual Meeting of the Association for Computational Linguistics, to be held in Uppsala Sweden. In 2009, together with James Curran from the University of Sydney, he led a team of 11 researchers at the Johns Hopkins University Language Engineering Workshop working on Large-Scale Syntactic Processing: Parsing the Web.