Research On Privacy Leak Detection System Based On Mobile Internet

Mass age has quietly approaching along with development and innovation of internettechnology and mobile communications technology. Mobile Internet provides rich andconvenient services for human, but also brings a huge security threat. Amount of personalprivacy information loaded by mobile terminal may be leaked through mobile internet.Privacy leak detection based on the mobile internet is an effective safety measure.Perfect privacy leak detection system should have the following characteristics: First,early warning for risk behavior. System should provide early warning for risk behavior thatmay result in the leak of privacy. Second, detecting leak of privacy. System should providetimely and effective detection when the user’s private information being leaked. Third,collecting new features of leak behavior. The lack of effective leakage characteristics is themain cause of failure of the early warning. So it is necessary to collect new features of leakbehaviors that can be used for early warning.Relative to the traditional terminal detection, this paper presents a detection idea forprivacy leak based on mobile internet after discussed security issues faced by the mobileinternet, the way of privacy leak and existing detection measures. Provide unified privacyleak detection for all users based on the data from mobile operator’s core network is thekey of this idea. The huge network traffic in the core network brings a great deal ofdifficulty to the real-time data acquisition, and then the diversity of the user’s privateinformation and the unpredictability of the network content enhance the difficulty ofdetecting leak of privacy. For these two aspects of the difficulties, this paper selected thecorresponding technical solutions according to the specific characteristics of the differentproblems and elaborated system design model and the implementation process based onthe research of data acquisition technology and pattern matching techniques. Due to theparticularity and restrictive of system deployment environment, this paper used virtualexperiment platform to validate and the result show that the system has high recognitionrate and detection efficiency.