The existing trajectory anonymity methods can not reflect the trajectory internal and external characteristics information well

and ignore personalized privacy requirements of moving objects.To solve these problems

we propose a new similarity measure model of trajectory structure

which considers the trajectory internal and external characteristics information of direction、speed、angle and location.On this basis

we propose the sparse minimum spanning tree clustering based personalized trajectory privacy protection algorithm.It reduces runtime by sparse methods

and generates an approximate optimal trajectory

k

-anonymity set by greedy strategy.Finally

the results showed our new similarity measure model of trajectory

structure can calculate distance of trajectories more accurately

and our method offers better utility and costs less time than previous proposals in the literature.