Research On P-Sensitive K Anonymity Privacy Protection Algorithm

With the advent of the era of big data,data release has received widespread attention.Data publishers publish the data they hold on the Internet,and various agencies make reasonable predictions on the data they obtain to improve social returns.On the one hand,the value of big data lies in its secondary use,so data publishers need to ensure that the availability of published data is maximized for usability.On the other hand,the data release will leak personal privacy information,and the data must be released on the premise of ensuring the security of the users' privacy information.Therefore,privacy protection based on the anonymous model has gradually become one of the research hot spots in data publishing.The commonly used anonymous models mainly include k anonymity,(?,k)anonymity,and(p,k)anonymity and so on.This paper mainly focuses on the(p,k)anonymity model,which presents a dynamic incremental update algorithm and a personalized anonymous algorithm,and applies them to the express information records.The main contents of the paper are as follows:(1)Incremental update algorithm for(p,k)anonymous model is proposed.Aiming at the shortcomings of the traditional(p,k)anonymous model,a data privacy protection algorithm for dynamic update is proposed.Combining encryption algorithms with(p,k)anonymous models to enhance the security of data to be updated.The interim table is used to store the data to be updated.The temporary table is used to record the categories of equivalence classes for published data,speeding up the speed of data partitioning equivalence classes.Posting updated data that meets the threshold and meets the(p,k)anonymity condition.(2)A personalized privacy protection algorithm based on(p,k)anonymous model is proposed.According to the users' perception of sensitive attributes,a personalized(p,k)anonymous algorithm is constructed.First,users are rated on the sensitivity of sensitive attributes sensitive attributes.Then,according to the score results,the sensitive attributes are divided into different levels and a generalized tree of sensitive attributes is established.Sensitive attributes with high sensitivity levels have arelatively high degree of generalization.Similarly,sensitive attributes with low sensitivity have a correspondingly low degree of generalization.Finally,an anonymous result that meets the need for personalized privacy protection can be obtained,and the degree of data protection can be improved while ensuring the availability of data.(3)An anonymous publishing privacy protection algorithm for express delivery information is proposed.In the era of online shopping,various types of e-commerce companies have a large number of customers' purchase records.Through the analysis of these records,they can discover potential business opportunities and provide a reliable strategy for the next marketing plan.Incremental update algorithm of(p,k)anonymity and personalized(p,k)anonymity algorithm are applied to express information of an electricity supplier,which not only meets the need for real-time updating of information,but also achieves the purpose of personalized privacy protection for customer express information.This paper deeply researches and analyzes the(p,k)anonymous privacy protection algorithm,proposes different improvements to the algorithm,and applies the improved algorithm to practical problems.Detailed experiments are conducted in conjunction with the actual collected data sets.The experimental and performance analysis shows that the improved algorithms presented in this paper have better performance and practical value.