AssociationRule on Vertically Partitioned Data
Poonam Lambhate, Rasika Khairnar · 2017
Data mining is process which extracts useful information of lager database of different applications like CRM, market basket analysis, web usage analysis etc. Cloud computing is very much powerful tool for storing data, which provides user to store their data remotely, same time data security must be preserved. So to provide data security to cloud data different techniques used such as encryption, cryptography etc. Each technique has its own pros and cons. In this paper we combine the association rules mining concept with data outsourcing procedure. D-Eclat algorithm is used for generating association rules. These generated association rules are securely outsourced on cloud server. Association rule mining (ARM) is applied on both horizontally and vertically partitioned data and proves that vertically partitioned outperforms in terms of time efficiency and system utilization. With the help of dynamic dataset, performance is measured and tested. Experimental results prove that system achieves better security and D-Eclat outperforms Eclat algorithm for rule generation.