期刊名称：Lecture Notes in Engineering and Computer Science
印刷版ISSN：2078-0958
电子版ISSN：2078-0966
出版年度：2018
卷号：2233&2234
页码：134-139
出版社：Newswood and International Association of Engineers
摘要：Modern cryptosystems, even though they give
confidentiality do not provide resilience against brute-force
attacks. Honey encryption (HE) was proposed as a
countermeasure to this problem of cryptography. HE is an
encryption scheme that supplies valid-looking, but fake
plaintext for every incorrect key used by an intruder to
decrypt a message. All possible messages relative to the
original message are mapped to a seed space such that any key
supplied by the attacker when decrypting a message produces
a relative, but fake message from the original message and this
makes it difficult for him to determine if he has recovered the
original message or not. However, a challenging problem with
HE is adapting it to natural language message to produce
convincing fake messages for documents such as emails. We
propose a novel approach of generating decoy message using
Stanford Dependency Parser and Wordnet from Princeton.
The result shows that the proposed scheme effectively produces
convincing decoy messages that fools the attacker. In addition,
the structure, length and content of the original message are
concealed. Finally, we verify the effectiveness of the proposed
scheme by checking the entropy of decoy messages from the
plaintext.
