Assessing the Disclosure of User Profile in Mobile-Aware Services

Mobile-aware services can be regarded as data-sharing systems in nature. In these systems, users obtain personalized service at the cost of sharing their personal information. As a result, it will inevitably lead to the disclosure of users' profiles and raise the serious privacy concerns. To assessing the privacy risk of sharing the user profile information items, in this paper we score and measure the potential risk of users caused by sharing information for the sake of personalization services. By adopted the 3-parameter logistic model, we explore information item's sensitivity, influence and probability of proper setting as well as users' potential attitudes to measure the privacy disclosure risk. The MMLE/EM algorithm is then adopted to estimate the above parameters. Finally, experiments on synthetic and real-world data sets are conducted and the results show that the obtained scores of our approach fit well with the real-world data.