CAJ | 학술논문

在基于空间事务的横向关联规则挖掘中,为了能够在海量数据中有效地提取空间拓扑关联规则,提出一种挖掘空间拓扑关联的有效算法,其适合挖掘多层横向空间关联规则.该算法用二进制数存储空间拓扑关系,使空间事务和数字建立对应关系,用数字递增的方法产生候选频繁项.在计算支持数时,算法在用逻辑运算的同时还利用数字特性减少扫描的空间事务数,大大地提高了效率.实验结果表明,在提取多层空间拓扑关联规则时,其比现有的算法更快速更有效.
재기우공간사무적횡향관련규칙알굴중,위료능구재해량수거중유효지제취공간탁복관련규칙,제출일충알굴공간탁복관련적유효산법,기괄합알굴다층횡향공간관련규칙.해산법용이진제수존저공간탁복관계,사공간사무화수자건립대응관계,용수자체증적방법산생후선빈번항.재계산지지수시,산법재용라집운산적동시환이용수자특성감소소묘적공간사무수,대대지제고료효솔.실험결과표명,재제취다층공간탁복관련규칙시,기비현유적산법경쾌속경유효.
In transverse association rules mining based on spatial transaction, to inefficiently extracte spatial topology association rules in large data, an eflicient algorithm of mining spatial topology association is presented, which is suitable for mining multilayer transverse spatial association rules. The algorithm uses binary number to save spatial topology association and founds a relationship between spatial transaction and number, and uses the way of number ascending to generate candidate frequent itemsets. Logic operation is used to compute support of candidate, and number character is used to reduce the number of scanned spatial transaction, which efficiently improve efficiency. The algorithm is used to extract multilayer spatial topology association rules, and this experiment indicates that the efficiency is faster and more efficient than presented algorithms.