Thesis Registration System

Login to register & upload your thesis: Log in

Kambhatla, Nishant - Decipherment of substitution ciphers with neural language models

  View the thesis

This thesis has been submitted to the Library for purposes of graduation, but needs to be audited for technical details related to publication in order to be approved for inclusion in the Library collection.

Term : Summer 2018

Degree : M.Sc.

Degree type : Thesis

Department : School of Computing Science

Faculty : Applied Sciences

Senior supervisor : Anoop Sarkar

Thesis title : Decipherment of substitution ciphers with neural language models

Given names : Nishant

Surname : Kambhatla

Abstract : The decipherment of homophonic substitution ciphers using language models (LMs) is a well-studied task in Natural Language Processing (NLP). Previous work in this topic score short local spans of possible plaintext decipherments using n-gram LMs. The most widely used technique is the use of beam search with n-gram LMs proposed by Nuhn et al. (2013). We propose a new approach on decipherment using a beam search algorithm that scores the entire candidate plaintext at each step with a neural LM. We augment beam search with a novel rest cost estimation that exploits the predictive power of a neural LM. This work, to our knowledge, is the first to use a large pre-trained neural language model for decipherment. Our neural decipherment approach outperforms the state-of-the-art n-gram based methods on many different ciphers. On challenging ciphers such as the Beale cipher our system reports significantly lower error rates with much smaller beam sizes.

Keywords : Natural Language Processing; decipherment; neural decipherment; neural language models; beam search

Total pages : 42