Secure Association Rule Mining Using Bi-Eclat Algorithm on Vertically Partitioned Databases

Association rule mining (ARM) is the important part of data mining, which helps to predict the association among multiple data items. The big challenge of ARM is efficiently extract the knowledge from large size databases of various applications. As per concern of data holder, the main challenge of ARM is to share the accurate information with protection of sensitive information. To achieve this, Privacy preserving ARM plays very important role. This paper presents the privacy preserving ARM over partitioned databases named as vertical partitioning of databases. In this, BiEclat algorithm is used to partition the database vertically and then identify the frequent item sets in all partition to mine the association rules. Further the research is enhanced by providing the security over mined association rules by using cryptographic techniques.