Privacy Preserving Approaches for Global Cycle Detections for Cyclic Association Rules in Distributed Databases
Nirali R. Nanavati, Devesh C. Jinwala · 2012
The current massive proliferation of data has led to collaborative data mining that requires preservation of individual privacy of the participants. A number of algorithms proposed till date in this scenario are limited to mining association rules and do not consider their cyclic nature that finds associations with respect to the time segment. Hence catering to this challenge, we propose techniques for privacy preservation while finding global cycles when mining cyclic association rules in a distributed setup. The proposed techniques are based on homomorphic encryption and Shamir’s secret sharing that can help us decipher partial and total global cycles along with maintaining privacy in a distributed setup. Additionally security, efficiency and correctness analysis of the proposed algorithms are also given.