Privacy Preserving Association Rule Mining in Partitioned Databases

Abstract: Fast improvement in the field of information technology rises the serious issue about huge information storage. To establish the correlation among set of data items, also becomes difficult challenge. To overcome these problems in huge data, association rule mining plays a very important role. In recent years, most of the researcher�s point of interest is to enhance the effectiveness of association rule mining based algorithm and thereby increasing the speed of mining. In existing system, association rules are applied on horizontal partitioned database and such technique is depending on IDMA, EMADS communication. In this paper, we consider both horizontal as well as vertical partitioned database, while applying association rule mining. The proposed system also minimizes the communications overhead. Further to enhance the security of system, RC4 algorithm is used to secure the horizontal and vertical partitions of database. Finally, system utilizes the protocol for recognizing the fake and duplicate frequent rules.

Keywords: Association Rule Mining, RC4, horizontal partitioned, vertical partitioned.