Research On Privacy Protection Of Location Based Services For Data Publishing

With the continuous progress of Internet technology,people's daily life has entered the era of big data,location big data service applications are constantly updated,user' trajectory data is constantly released,through analysis and mining,researchers can continuously provide users with more convenient services,such as personalized recommendation of interest points,social space sharing of friends,and so on.However,the publication of trajectory data usually contains the user's personal sensitive information,and the user also bears the risk of personal privacy disclosure while enjoying the convenience of location service.For example,the attacker can infer the physical health status,family information and so on according to the location point in the user's trajectory,and infer the personality preferences and economic situation by querying the user's access frequency to certain points of interest.Therefore,how to reduce the risk of user privacy disclosure while ensuring the availability of trajectory data has become an important subject for people to study deeply.According to the above,the following research is carried out in this paper.(1)This paper expounds the necessity of trajectory data in location-based services,introduces the measurement standard of privacy protection,analyzes and summarizes the current commonly used trajectory data privacy protection methods,and introduces the definition and implementation mechanism of differential privacy and the attack methods based on background knowledge.(2)In order to prevent the data loss phenomenon of static trajectory data in the publishing process of privacy protection,a semantic generalization trajectory data publishing algorithm based on information gain is proposed in this paper to solve the problems of improper generalization and resource waste caused by classification errors in the current generalization data publishing mechanism.The main contents of the algorithm include:introducing the information gain availability function into the trajectory data publishing;using the information gain availability function to accurately determine the split nodes of the semantic generalization tree and reasonably establish the semantic generalization tree;reasonably setting the semantic generalization distance according to the user's different requirements for privacy protection to achieve semantic generalization to protect the user's privacy information security.It is verified by experiments,that the algorithm can effectively reduce the data distortion caused by improper classification generalization,and meet the requirements of improving user privacy security and data availability.(3)In order to solve the phenomenon of user privacy leakage in the publishing process of dynamic trajectory data privacy protection,a differential privacy trajectory data publishing algorithm based on local suppression is proposed in this paper,which can better solve the problem of data loss and low privacy security in the current suppression data privacy protection publishing method.The main contents of the algorithm include: in the process of data processing,the differential privacy protection mechanism is combined;in the process of local suppression,the minimum violation sequence set is constantly updated by calculating the priority score of the inhibition points,so as to meet the LKC-privacy model and reduce the loss rate of frequent sequences;and establish the classification tree according to the updated trajectory data set,allocate the privacy budget reasonably and add noise to the data to ensure that the privacy of the data.It is verified by experiments that the algorithm can greatly reduce the sequence loss caused by excessive suppression,ensure that the published data has high availability value,and effectively improve the security of trajectory data after publishing.