Research On The Privacy Protection Technology Of The Numerical Sensitive Attributes Based On Dynamic Datasets

With the social progress and progressive development of computer technology, privacy protection technology has captured a lot of people's concern. Lots of personal information is stored in departments such as the government, hospitals and banks, some information is needed to be distributed on the Internet. Data releasing is seen as a means of resource sharing and it provides great convenience for data exchange and information sharing. At the meantime, it would also result in the leakage of sensitive personal information. At present, the existing anonymous distribution technologies are mainly based on static datasets and only one distribution is supported in the datasets one time. However, data in the real world are changing at all times. Therefore, how to anonymize dynamic datasets has caused concern among many researchers.Since numerical sensitive attribute privacy protection technologies in the static environment can not meet the needs of people, incremental numerical sensitive attribute privacy protection method was proposed in the dynamic environment. First, by comparing the multiple distribution of incremental datasets, a comprehensive analysis of incremental value sensitive attribute redistribution that would lead to the risk of potential privacy leakage was conducted. Second, in terms of the characteristics that numerical sensitive attribution would result in proximity attack and incremental value sensitive attribute redistribution would cause privacy leakage. On the basis of generalization thought, we put forward effective algorithm on anonymous method against republication of incremental numerical sensitive data. When the record increasing, if it meets the condition, then the newly increased records and original records will be distributed together. If it does not meet the anonymous condition, the records will be redistributed with the utilization of bucket algorithm and put into g small buckets. Then the new records and original anonymous records will be distributed again. The method makes sure that proximity attack doesn't occur between the numerical sensitive attributes, and realizes incremental numerical sensitive attributes anonymous republication.Finally, we test the incremental numerical sensitive attributes republication anonymous method through a lot of experiments, these experiment results show that this method can improve the privacy protection degree and running time efficiency, which makes sure that privacy information will not be leaked.