CAJ | 학술논문

指出了时序关联规则在支持度度量上的不足，存在处理高密度海量数据时往往要耗费大量的时间处理规模巨大的频繁候选集，同时需要多次重复扫描数据库，执行效率低等问题，提出了重新定义支持度度量方法的时序关联规则。针对时序关联规则在解决项分类时涉及到的分层不确定、不准确情形，引入了基于隶属度的模糊层次分类结构，定义了项间距离、项集间距离，最终得到一种新的关联规则间距离的度量方法。实现了模糊层次分类，将时序关联规则结果进行聚类分析，得到规则和规则之间相似性，实验验证了该方法的有效性。
지출료시서관련규칙재지지도도량상적불족，존재처리고밀도해량수거시왕왕요모비대량적시간처리규모거대적빈번후선집，동시수요다차중복소묘수거고，집행효솔저등문제，제출료중신정의지지도도량방법적시서관련규칙。침대시서관련규칙재해결항분류시섭급도적분층불학정、불준학정형，인입료기우대속도적모호층차분류결구，정의료항간거리、항집간거리，최종득도일충신적관련규칙간거리적도량방법。실현료모호층차분류，장시서관련규칙결과진행취류분석，득도규칙화규칙지간상사성，실험험증료해방법적유효성。
This paper firstly indicates the problems of time-serial association rules in measuring support and classifying items,which costs a lot of time to deal with a large scale candidate set of frequent items when processing mass data of high density and needs to repeatedly scan the database with low execution efficiency .To avoid the vagueness and inaccuracy of classification, fuzzy hierarchical classification structure based on membership degree is introduced ,and the distance between items and the distance between item sets have been defined ,consequeatly , a new method to measure the distance between association rules has been given .The proposed method achieves fuzzy hierarchy classification , conducts clustering analysis of time-serial association rules ,and results in similarity between rules and rules .The effectiveness of the proposed method is verified by experiments .A new method for the previous time -serial association rules has been redefined in measuring support .