Abstract:
In today digital age, there are more
demands to improve techniques for information
security. One of the solutions is steganography,
hiding sensitive information within innocentlooking
cover media. In this paper, we propose
a syntax-based steganography approach by
using the syntax bank. The input raw sentence is
parsed with the Stanford parser that can
produce a phrase structure of the sentence. This
parse structure is then used to produce the
syntax of the sentence. At the same time,
Shannon-Fano coding is used to compress the
input secret message as minimum total bits
length as possible. Then, syntax transformation
task searches the syntax set of the given
sentence within the syntax bank, and then
transforms it into a desired syntax that can
represent the key-controlled semi-randomly
generated secret bits intended to hide in the
sentence. The resulting stego text will still be
innocent-looking by applying semantically
unchanged syntax transformation on the input
text. Thus, the detection of the secret message
may be hard for the intruder.
