一种基于投影FP-growth的co-location模式挖掘算法  

A Project FP-growth-based Algorithm for Mining Spatial co-location Patterns

作　　者：余翠兰 

机构地区：[1]云南德宏师范高等专科学校计算机科学系,潞西678400

出　　处：《科学技术与工程》2014年第23期234-240,共7页Science Technology and Engineering

基　　金：云南省教育厅科学研究基金项目(2013Y571)资助

摘　　要：空间co-location模式代表了一组空间属性的子集,它们的实例在地理空间中频繁地关联。针对如何利用关联规则挖掘算法来快速地挖掘co-location模式的问题,通过采用FP-CM算法与投影频繁模式树(PFP_tree)及其他技术相结合的方法,提出了一种基于投影FP-growth的co-location挖掘算法,简称PFP_CM算法。这个新算法主要对产生最大频繁模式的方法、模式过滤的方法、访问数据库的次数、避免大量的表实例连接操作的方法进行改进。最后通过大量的实验,验证了该算法的高效性和正确性,同时,将其用于对三江并流地区珍稀植物的共生物种进行挖掘。The specification Spatial co-location patterns are traditionally defined as the subsets of features whose instances are frequently located together in geographic space. A novel approach for mining co-location patterns is proposed, which is called PFP_CM （the Project FP-growth based Co-location Miner）. This new approach is FPCM combined with Project FP tree and other technologys. It is presented that how to use the association rule mining algorithm for mining co-location patterns. This new algorithm is improved in the methods of generating maximum frequent patterns, filtering patterns, scanning database and voiding plenty of table instance joins. Finally, by extensive experiments, the effectiveness and correctness of our algorithms are verified. They are used in a plant distributing dataset of ＂Three Parallel Rivers of Yunnan Protected Area＂ to mining symbiotic plant species,

关 键 词：空间数据挖掘 同位模式 关联规则 最大频繁模式 投影频繁模式树 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]