System Construction by Content-Based Classification on Metadata by Classification Adaptive Policy Prediction

Abstract

With the incrementing volume of images users share through convivial sites, maintaining privacy has become a major quandary, as demonstrated by a recent wave of publicized incidents where users inadvertently shared personal information. In light of these incidents, the desideratum of implements to avail users control access to their shared content is ostensible. Toward addressing this need, we propose an Adaptive Privacy Policy Prognostication (A3P) system to avail users compose privacy settings for their images. We examine the role of convivial context, image content, and metadata as possible designators of users’ privacy predilections. We propose a two-level framework which according to the user’s available history on the site, determines the best available privacy policy for the user’s images being uploaded. Our solution relies on an image relegation framework for image categories which may be associated with kindred policies, and on a policy presage algorithm to automatically engender a policy for each incipiently uploaded image, additionally according to users’ convivial features. Over time, the engendered policies will follow the evolution of users’ privacy posture. We provide the results of our extensive evaluation over 5,000 policies, which demonstrate the efficacy of our system, with presage accuracies over 90 percent.