基于有序FP-tree结构和投影数据库的最大频繁模式挖掘算法  

Maximal Frequent Pattern Mining Algorithm based on Ordered FP-tree Structure and Projection Database

作　　者：王利军[1] 唐立 WANG Li-jun;TANG Li(Department of Information Engineering,Anhui Institute of Economics Management,Hefei Anhui 230031,China)

机构地区：[1]安徽经济管理学院信息工程系,安徽合肥230031

出　　处：《淮阴师范学院学报（自然科学版）》2020年第1期35-39,44,共6页Journal of Huaiyin Teachers College;Natural Science Edition

基　　金：安徽省高校自然科学重点资助项目(KJ2019A0965);安徽经济管理学院教学研究资助项目(yjjyxm201903)。

摘　　要：OPFP-MAX是基于有序FP-tree结构和投影数据库的最大频繁模式挖掘算法。有序FP-tree结构不仅可以减少空间的浪费,而且由于树结构的有序性,在挖掘数据时可以减少挖掘事务项的数量,从而加快挖掘效率。算法采用垂直投影方案实现数据分解,基于投影数据库创建局部树结构,从而进一步减少对系统空间资源的浪费。采用优化策略加快挖掘效率。实验验证了算法的可行性和优越性。OPFP-MAX was a maximal frequent pattern mining algorithm based on ordered FP-tree structure and projection database.Ordered FP-tree structure can not only reduce the waste of space,but also reduce the number of transaction items when mining data due to the orderliness of tree structure,thus speeding up the mining efficiency.The vertical projection scheme is used to implement data decomposition,and the local tree structure is created based on the projection database,which further reduces the waste of system space resources.The optimization strategy is adopted to speed up the mining efficiency.Experiments verify the feasibility and superiority of the algorithm.

关 键 词：有序FP-tree 投影数据库 最大频繁项集 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]