Apriori Parallel Improved Algorithm Based on MapReduce Distributed Architecture

摘要：

　　Under the environment of big data,efficiency is low and there are many candidates when the traditional serial Apriori algorithm in dealing with massive data.This paper proposes a parallel better algorithm based on MapReduce distributed architecture.Based on the basic Apriori algorithm on MapReduce,this paper makes a reconstruction of the original transaction database,and implements parallel in data set fragmentation.The algorithm optimizes the transaction database; candidate item sets counting and pruning strategy.The experimental results show that the improved algorithm proposed in this paper can reduce the candidate items and improve the efficiency.

关键词： Apriori algorithm MapReduce distributed

作者: She Xiangyang Zhang Ling

作者单位: College of Computer Science and Technology Xian University of Science and Technology Xian, China

会议类型: 国际会议

会议名称: 2016 Sixth International Conference on Instrumentation and Measurement,Computer,Communication and Control (IMCCC2016)(第六届仪器测量、计算机通信与控制国际会议)

会议地点: 哈尔滨

会议语种:英文

页码: 517-521

在线出版日期: 2016-07-21（万方平台首次上网日期，不代表论文的发表时间）