A quantifiable privacy protection method, device and medium for destination prediction

Abstract

The invention discloses a quantifiable privacy protection method, device and medium for destination prediction, wherein the method is: using the Laplacian mechanism to inject noises of different radii into multiple sets of the same historical trajectory data sets to obtain corresponding trajectories Data set; predict the destination of each set of trajectory data sets, and calculate the corresponding degree of privacy protection according to the prediction results to construct training samples, and use the noise radius level as the label value of the training sample; use all training samples and their label values ​​to train A privacy quantitative protection model based on multiple linear regression; when receiving the privacy protection degree requirements and the track to be protected, input the privacy protection degree requirements into the privacy quantitative protection model; use the Laplacian mechanism to output the privacy protection model according to Noise radius level to inject noise into the track to be protected. The invention can fully meet the user's privacy requirement, and provide the user with accurate and stable track privacy protection.

Description

technical field [0001] The invention belongs to the field of privacy protection and information security, and aims at location privacy threats of high-accuracy destination prediction services, and proposes a quantifiable privacy protection method, device and medium for destination prediction. Background technique [0002] With the popularity of mobile devices with GPS and the continuous development of Internet of Things technology, more and more location-based services facilitate all aspects of our lives. Such as business activities, personal health management, etc. In recent years, a technique for predicting a user's destination has been developed. This technology predicts the location that the mobile user wants to reach based on the partial trajectory of the mobile device user and the prediction model established by the historical trajectory database of the city. This technology has significantly promoted the development of location-based services, but it also brings gre...

AI Technical Summary

Problems solved by technology

However, this method is too rough and provides extremely unstable privacy protection, and there are some special cases (for example, there are stops and other positions at the beginning and end) that make the protection method invalid

Images