Reconstruction Based Privacy Preservation in Centralized Incremental Data Mining

The technological development has led the storage of data efficiently in terms of both storage and cost. This huge amount of data can be used for research to reveal the hidden information for betterment of life. Since the data may hinder the privacy of individual, there is an emerging research area going on towards privacy preservation of data. When many are concentrating on static data, this is a work developed solely for incremental data where in the database gets updated frequently. The method groups the data into various classes and the encryption is based on the key values generated within each class. Since the key is not a constant private or public key, the method provides a greater amount of protection compared to usual cryptographic techniques. In this paper RPPCID, the algorithm is specified with a sample input and output database. The method does not require the execution of entire database after insertion. The method is a combination of cryptography and perturbation methodology and hence has the advantages of both. The performance of the methodology is also expressed using a graph.