Steganography in Social Networks Based on Behavioral Correlation

Yinghong Hu,Zichi Wang,Xinpeng Zhang
DOI: https://doi.org/10.1080/02564602.2020.1721340
2020-01-01
Abstract:This paper proposes a secure steganographic method in social networks based on the behavioral correlation between the sender and his friends. A sender transmits secret data to one of his friends in social network by marking "love" on the news published by his friends. To avoid behavioral abnormality, the sender lets the value of "love" marking probability on each news in proportion to the number of "love" marks made by his friends. As a result, the "love" marks carrying secret data are indistinguishable from other normal "love" marks. Although a receiver does not know the "love" marking probabilities, he can still extract the secret data by observing the "love" marks. The security performance of the proposed steganographic method is satisfactory when examined by a proposed behavioral steganalysis method which is able to defeat existing behavioral steganographic methods.
What problem does this paper attempt to address?