An Anonymous Rules With Application Of Rough Set Theory And Research Of Microaggreation Algorithm

The fast developing Information Techniques impact every walk of life deeply where large scales of data has been collected. In the data, lots of useful information is implicated.In the process of mining useful information from these data, some personal information may be published. And some may contain the privacy of certain people if data without disposing is published. Therefore, privacy invasion happens, the problem of finding a method for publishing the useful data which can also preserving privacy remains to be solved.The method of publishing with restriction works at the cost of information loss (Publishing the data with lower accuracy when comparing with the original data) Studies on these methods always focus on anonymous ways.Anonymity of data has been widely studied for the variety of data types it can process and the acceptable information loss level during the process of data publishing.Other merits like high level of privacy invasion protection, common computing cost, acceptable information loss, low dependence from data and so on also facilitate it become the focus for privacy preserving.For the conventional anonymity algorithms, the great loss of information when processing data of high dimension renders its practicability low. Microaggregation algorithm adopt the thinking of clustering, is effective as well as simple. Especially for continuous data, such algorithm has the particular predominance of holding more information about semanteme.This paper designs a new anonymous rules and achieves a Microaggregation algorithm which is able to generalize mixed data. And for categorical data, the application of rough set theory, proposed an assessment of data availability model.With the comparison test in the universal data set:in order to ensure privacy protection in the same degree of improvement, this method has less information loss than classical anonymous rules, namely realizes effectively in order to increase the availability of data privacy foundation.