Privacy protection method for sparse position trajectory

Abstract

The invention discloses a privacy protection method for a sparse position trajectory, a position point candidate set is obtained by dynamically adjusting an anonymity threshold, the dynamic position anonymity threshold can reduce inhibition and reduce a trajectory point inhibition rate compared with a static position anonymity threshold, and in addition, historical data of a user is not adopted asthe basis of a false trajectory. Direct exposure of data is avoided; and finally, connecting the position points in the position point candidate set by judging the accessibility between the adjacentposition points and combining the local transition probability to obtain a candidate track set. And similarity judgment is performed between the tracks in the candidate track set and the original track, and the optimal k-1 tracks are selected from the candidate track set in combination with the overall transition probability, so that the analysis on the transition probability can be effectively resisted, and the anonymous effect is greatly improved.

Description

technical field [0001] The present invention relates to the technical field of privacy protection, in particular to a privacy protection method for sparse location trajectories. Background technique [0002] Traditional trajectory privacy protection schemes often generate a set of anonymous trajectories. Before the trajectory data is released, some sensitive locations in the trajectory are deleted. This method is called trajectory suppression. The sensitive location here refers to the sparse geographic location in the trajectory. When there are too many suppression points, the availability of data release will be poor, and if there are too few suppression points, the privacy protection effect will be poor. Therefore, how to reasonably suppress location information is an important problem currently facing. [0003] In the existing technology, Yannian Zhao et al. can analyze and attack from the starting point and end point of the user's trajectory based on the background knowl...

AI Technical Summary

Problems solved by technology

[0004] However, selecting location points and trajectory segments from the user's historical trajectory to fit a false trajectory requires a large amount of personal data of the user. When the data is insufficient, it is difficult to achieve a high privacy protection effect, and even the historical data set is the user's real information, there is also a risk of leakage

What's more, when using the trajectory suppression method, there is no reasonable suppression. When most of the user's trajectory points are in a sparse environment, there is too much suppression, resulting in poor availability of published data

In addition, when using randomly generated location points on the map, geographical features are not considered. The generated location points may be in places that ordinary people cannot reach. When fitting the trajectory, the transition probability between local locations and the overall trajectory transfer are not considered. probability

Images