A fast distributed algorithm for mining association rules

Abstract

Abstract

A SURVEY-EFFICIENT & SECURE MINING FOR VERTICAL DISTRIBUTED DATABASE

“…Kantarcioglu and Clifton in [13] use a secure multiparty computation to model the horizontal partitioning of transactions across sites, and present algorithms that incorporate cryptographic techniques to minimize the shared information without incurring much overhead in the mining process. The paper in [14] proposes an efficient distributed algorithm FDM (Fast Distributed Mining of association rules) for mining association rules. Some interesting properties between local and global large itemsets are observed, which leads to an effective technique for the reduction of candidate sets in the discovery of largeitemsets.…”

Association Rule Mining using Apriori Algorithm for Distributed System: a Survey

“…Apriori algorithm was designed to run on databases containing transactions. It is a "bottom up" approach as candidate items are first generated and then the database is scanned to count the support for candidate items exceeding minimum support [23]. The number of items in candidate subsets is increased one at a time with iteration.…”

“…Concept of parallel processing [22] and distributed computing [23] are helpful to solve above problems so we can combine both the concepts. We can divide the data then generate processing node and then we can process that distributed data separately in each node after processing we combine the results.…”

Research on distributed data mining system and algorithm based on multi-agent /

“…and DHP [66] put forward by Park, etc., and the dividing algorithm PARTITION [67] put forward by Savasere, etc., and the sampling [68] algorithm put forward by Tovionen, etc., and some updating algorithms of the association rule such as FUP, IUA and NEWIUA, and so on. The third one is the algorithm solving the problems of the association rule mining in the distributed environment, such as DMA [71] , FDM, etc. [7278] .…”