Personalized Privacy Protection With Spatio-temporal Features In Social Networks

With the development of mobile services in social network,the service function in social application collects a lot of personal information.Although these service functions provide great convenience for users' life,protecting users' privacy information has become a key issue.In reality,users have different privacy protection requirements,and attack protection technologies with the same background knowledge are insufficient to meet users' personalized privacy protection requirements.Therefore,a complete privacy protection model is needed to solve the problem that the leakage of users' personal information in the process of publishing and sharing the data with spatio-temporal features.First of all,the paper standardized the definition of personalized privacy protection with spatio-temporal features in social networks,elaborates the data privacy derivation mechanism in detail and proposes to protect the sensitive attributes of the data with spatio-temporal features by using personalized anonymous technology.According to the extension of the data with spatio-temporal features as attackers' background knowledge,the data with spatio-temporal features identification is defined as four levels by different sensitive attributes,and laid the foundation for the design and implementation for personalized data privacy protection technology.Secondly,the personalized data privacy protection algorithm is proposed,the implementation mechanism of the four levels of protection methods in the algorithm is explained.By means of data transformations to generalize the data with spatio-temporal features to the same area in the same period,the attacker cannot get the users' privacy information from the data with the same quasi-identifier.The users could set the personalized protection level,in this situation,they could set the corresponding constraints on sensitive attributes and perfect the users ' personalized privacy protection requirements at the same time.Finally,the experiments are performed based on real social network data,and comparison and analysis are performed.The experimental results show that the personalized privacy protection algorithm oriented to spatio-temporal features could effectively protect the users' data privacy while ensuring the availability of the data with spatio-temporal features.