Detecting Inference Channels in Private Multimedia Data via Social Networks

Indirect access to protected information has been one of the key challenges facing the international community for the last decade. Providing techniques to control direct access to sensitive information remain insufficient against inference channels established when legitimate data reveal classified facts hidden from unauthorized users. Several techniques have been proposed in the literature to meet indirect access prevention. However, those addressing the inference problem when involving multimedia objects (images, audio, video, etc.) remain few and hold several drawbacks. In essence, the complex structure of multimedia objects makes the fact of detecting indirect access a difficult task. In this paper, we propose a novel approach to detect possible inference channels established between multimedia objects representing persons by combining social network information with unmasked content of multimedia objects. Here, we present the techniques used to map the content of social networks to the set of multimedia objects at hand. We also provide an MiD function able to determine whether an unmasked multimedia object combined with data from the social network infers a sensitive multimedia object.