Securing Text Steganography Based on Synonym Substitution Against Word Embedding Steganalysis

HENDRO PRASETYO

Informasi Dasar

23.05.052
004.071
Karya Ilmiah - Thesis (S2) - Reference

Steganalysis is a method to attack the steganography system. One of steganalysis is based on synonym substitution. This steganalysis attack is conducted by analyzing the context fitness value of the sentence. A steganography system based on synonym substitution by setting the value of context fitness of the cover text not at the maximum value is proposed. Using this method, the steganalysis cannot distinguish between the cover and stego text. For realizing the proposed steganography method, five processes are necessary. They are context fitness for feature detection, context-free grammar to create cover text, Quantum Random Number Generator to generate binary random values used to modify the cover text, Zero-Width Characters to hide the binary key, and T-Lex System for the message embedding. The steganalysis accuracy using the proposed method ranged from 5.58 % to 16.00 %. Meanwhile, using the previous method, the steganalysis accuracy ranged from 86.41% to 97.91%. This evidence proves that word embedding-based steganalysis (WES) cannot work effectively on the proposed steganography method.

Subjek

JOURNAL OF INFORMATICS AND COMPUTER SCIENCE
STEGAMOGRAPHY,

Katalog

Securing Text Steganography Based on Synonym Substitution Against Word Embedding Steganalysis
 
 
Inggris

Sirkulasi

Rp. 0
Rp. 0
Tidak

Pengarang

HENDRO PRASETYO
Perorangan
Ari Moesriami Barmawi
 

Penerbit

Universitas Telkom, S2 Informatika
Bandung
2023

Koleksi

Kompetensi

  • CSH623 - TESIS
  • CSH6I3 - TOPIK KHUSUS DALAM PENYEMBUNYIAN INFORMASI PADA TEKS, CITRA DAN VIDEO
  • CSG6F3 - TOPIK KHUSUS DALAM STEGANALISIS

Download / Flippingbook

 

Ulasan

Belum ada ulasan yang diberikan
anda harus sign-in untuk memberikan ulasan ke katalog ini