期刊名称：International Journal of Engineering and Computer Science
印刷版ISSN：2319-7242
出版年度：2015
卷号：4
期号：3
页码：10924-10929
出版社：IJECS
摘要：We proposed a protocol for protected mining of association rules in horizontally scattered database. Thecurrent leading set of rules is that of Kantarcioglu and Clifton. Our procedure, like theirs, is based on theFast Distributed Mining (FDM) algorithm of Cheung et al. which is an unsecured spread version of theApriori algorithm. The main ingredient in our procedure are two novel secure multi-party algorithms[1] —one that computes the union of private subsets that each of the interacting group of actors hold, and anotherthat tests the inclusion of an element held by one actor in a subset held by another. Our protocol offersimproved separation with respect to the protocol. In addition, it is simpler and is extensively more efficientin terms of announcement rounds, announcement cost and computational cost
关键词：Privacy Preserving Data Mining;Distributed Computation; Frequent Itemsets;Association Rules.
