Research On Anonymity Privacy Preserving In Data Publishing

With the development of information technology,personal information is widely collected,published and used.The efficient circulation and use of these data has greatly facilitated people's daily life.However,usually,there are a lot of personal privacy information are hidden in these data.Unreasonable use of these data may lead to leakage of personal privacy.Therefore,how to protect personal privacy,ensure data security and prevent sensitive information leakage has become an urgent issue in the field of information security.Anonymity privacy preserving technology is an important method to protect the security of personal privacy.It ensures the security of sensitive information while considering the validity of data.Achiving an balance between information security and effectiveness of data.According to the actual demand of privacy preserving and the shortage of the existing static anonymit models,this paper propose a static anonymity privacy preserving model based on sensitivity.Aiming at the problem of over generalization in the existing static anonymity models,this paper introduces the idea of clustering analysis into static anonymity models.Using the distance calculation method in clustering to calculate the distance between the tuples,minimizes the difference of the tuples in the same equivalent class and maximizes the difference of the tuples between the different equivalent classes.The corresponding distance calculation method and clustering generalization method are given for different types of attributes.In view of the problem that the existing static anonymity models do not consider the sensitivity.In combination with the demands in privacy preserving that the sensitive attributes of different sensitivity should be provided with different intensity of protection.We introduce sensitivity into anonymity preserving,and give the sensivity measurement method.According to the sensitivity of sensitive attributes,sensitive attribute values are grouped into different groups.Then,set the corresponding strength constraints for different sensitive attribute groups.Experimental results show that the static anonymity model proposed in this paper can effectively protect user's privacy data and ensure the availability of data.In existing dynamic anonymous models,most can only process dynamic data sets that only include insert and delete updates.This paper takes the attribute value updates into account,and proposes a dynamic anonymity model that can handle complete dynamic data sets(including insert,delete,and attribute value updates).We analyze different dynamic updates,and define dynamic data sets according to the different updates existing in data sets.It also analyzes the change of sensitivity when the sensitive attribute values are dynamically updated.For different dynamic updates of sensitivity,putting forward corresponding protection strategie.Experimental verification shows that the dynamic anonymity model proposed in this paper can effectively protect the fully dynamic data set,and the availability of data after anonymity processing is high.In this paper,we propose a static anonymity model and a dynamic anonymity model for static and dynamic datasets.It provides a new solution for privacy preserving in data publishing.Theoretically,it enriches the anonymity privacy preserving models,and improves the anonymity preserving research based on sensitivity.In the practical sense,it increases the flexibility of privacy preserving strategy,which not only ensures data privacy and security,but also improves data availability.