Research on Trajectory Clustering Optimization Algorithm Based on Sparse Representation

Privacy issues in the trajectory play an significant role with the popularization of intelligent terminals in the age of big data. Personal information about users is easily exposed to the attackers in LBS. A big number of solutions are presented in the literature to better protect privacy. The scheme based on k-anonymity model has been widely used. However, the traditional methods have some limitations due to their long convergence time and subjective factors. In this paper, we propose a scheme based on sparse representation, which has high computational efficiency and simple solution. Firstly, we process the data set with certain techniques. Then we introduce the expression of sparse representation between trajectories and add regularization for continuous training. Experiments show that the optimization model achieves better clustering results, and the validity of the algorithm is testified by different evaluation indexes.