Online Social Networking Sites attracted a massive number of users over the past decade but also raised privacy concerns with the amount of personal information disclosed. Studies have shown that 25% of the users are not aware of privacy settings provided by these sites or do not know how to change them. This paper investigates an approach towards understanding users' privacy behavior on social media, e.g. Facebook, through studying faces, tags and photo privacy settings. It classifies users based on their privacy selections and proposes a system for monitoring and recommending stronger privacy settings. An application is developed, and our case study examines the effectiveness of our model.