分布式數(shù)據(jù)挖掘計(jì)算過(guò)程-DDCP算法研究

提出了一種關(guān)聯(lián)規(guī)則挖掘大項(xiàng)集生成的并行和分布式處理的計(jì)算框架的算法，該算法以大規(guī)模事務(wù)數(shù)據(jù)庫(kù)為基礎(chǔ)，將數(shù)據(jù)有效地分片后作分布或者并行處理，通過(guò)節(jié)點(diǎn)之間的通信降低了節(jié)點(diǎn)間傳輸?shù)臄?shù)據(jù)量。通過(guò)算法實(shí)例驗(yàn)證了算法的正確性和可行性，可以在分布式或者并行環(huán)境里實(shí)現(xiàn)高效的數(shù)據(jù)挖掘。
關(guān) 鍵 詞 數(shù)據(jù)挖掘; 關(guān)聯(lián)規(guī)則; 大項(xiàng)集; 數(shù)據(jù)庫(kù)

Abstract This article proposed a algorithm of the calculate architecture used for the association rule and this algorithm based on the data partition, fully uses the merits and specialties, at the same time uses controller to assign transactions randomly to resolve the data skew in the database. The algorithm is used for the example and shows the correctness and feasibility．It can be used for distribute database and most applicable for distribute calculation.
Key words datamining; association rule; large itemset; database