Privacy and Efficiency Aware User-Generated Content Access in Social Medi

Social media has boomed the creation and propagation of user-generated contents (UGCs), which brings new opportunities for widespread applications. To guarantee the user experience of accessing the UGCs in social media, we are faced with two dilemmas. First, the easy access of UGCs for various social media users threatens the privacy of the UGC owners, whereas privacy preservation may degrade the extent of user interactions. Second, the considerable amount of UGCs require abundant bandwidth and storage resources for easy access, whereas the increase of available resources always lags behind the in- crease of UGC data.;In this dissertation, we propose a suit of new designs to tackle the privacy threatens and efficiency issues in accessing UGCs in social media. The dissertation can be schematized in three parts.;First, we address the privacy preservation problem in social media sharing. Traditional access control mechanisms, where a single access policy is made for a specific piece of content, cannot satisfy the user privacy requirements in large-scale media sharing systems. Instead, configuring multiple levels of access privileges for the shared media contents is desired to facilitate the widespread propagation of media contents, and to accord with the diverse and complex social relationship among users. In this research, we propose a scalable media access control (SMAC) system to enable such a configuration in a secure and efficient manner. The proposed SMAC system is empowered by the scalable ciphertext policy attribute-based encryption (SCP-ABE) algorithm as well as a comprehensive key management scheme. We provide formal security proof to prove the security of the proposed SMAC system. Additionally, we conduct intensive experiments on mobile devices to demonstrate its efficiency.;Furthermore, we study the privacy aware location-based services (PA-LBSs), which preserve LBS users' privacy but undesirably sacrifices the service quality. In order to balance the two factors with satisfactory user experience, existing frameworks are faced with two barriers, i.e., scalability and social-friendliness. First, existing schemes do not provide flexible privacy protection mechanism that enables LBS users to choose privacy levels scalably. The lack of scalability easily results in either unacceptable service quality degradation or insufficient privacy protection and fails to meet the dynamic user requirements. Second, existing schemes handle privacy protection by merely considering the trust relationship between users and servers but ignore the complex trust relationships among users. As a result, users cannot preserve privacy in location-based social services that involve user-to-user interactions. In this research, we present the first scalable and social-friendly PA-LBS (SSPA-LBS) system. In particular, we propose a novel camouflage algorithm with formal privacy guarantee that enables LBS users to scalably expose their location information from two key perspectives. Furthermore, we apply the SCP-ABE algorithm to enable LBS users to effectively control the access from other users to their location information. Moreover, we also demonstrated the operational efficiency of the proposed system through successful implementations on Android devices.;In the third part of this dissertation, we study the efficient access of user generated videos (UGVs) that consume the largest network resources among social media applications, by leveraging the ability of video popularity prediction. While using historical popularity can predict the near-term popularity with a reasonable accuracy, the bursty nature of online content popularity evolution makes it difficult to capture the correlation between historical data and future data in the long term. Although various existing efforts have been made toward long-term prediction, they need to accumulate a long enough historical data before the prediction and their model assumptions cannot be applied to the complex networks with inherent unpredictability. In this research, we propose LARM, a lifetime aware regression model, representing the first work that leverages content lifetime to compensate the insufficiency of historical data without assumptions of network structure. The proposed LARM is empowered by a lifetime metric that is both predictable via early-accessible features and adapt- able to different observation intervals, as well as a set of specialized regression models to handle different classes of videos with different lifetime. We validate LARM on two YouTube data sets with hourly and daily observation intervals. Experimental results indicate that LARM outperforms several non-trivial baselines from the literature by up to 20% and 18% of prediction error reduction in the two data sets.