摘要
关联规则研究数据库中一组互不相属对象之间的相关性,挖掘出具有一定意义的关联关系、挖掘算法如Apriori、FP-Growth等,这些算法需要反复多次扫描整个数据库导致I/O负载增加,降低了CPU的性能.文章通过对数据库进行转置和平行变换以减少扫描的次数,从而提高算法效率.
Association rules are interesting correlations among attributes in a database.It plays an important role in generating frequent item set from large database.It is discovered that interesting association relationship is among business transaction records in many business decision-making process,such as catalog decision,cross-marketing,and loss-leader analysis.It is also used to extract hidden knowledge from large datasets.The Association Rules Mining algorithms such as Apriori,FP-Growth require repeated scans over the entire database.All the input/output overheads are being generated during repeated scanning of the entire database and decreases the performance of CPU,memory and I/O overheads.The Algorithm uses databases in transposed form and database transposition is done using Parallel transposition algorithm so as to generate all significant association rules number of passes required reduced.
作者
马良斋
周渊
姜滨
Ma Liang-zhai Zhou Yuan Jiang Bing(School of Electronic and Information Engineering, Lanzhou Jiaotong University School of Chemical and Biological Engineering, Lanzhou Jiaotong University,Lanzhou Gansu 730070)
出处
《河西学院学报》
2016年第5期46-51,共6页
Journal of Hexi University
基金
甘肃省自然科学基金项目(项目编号:1212RJZA059)
