基于等价关系的关联规则的挖掘Equivalence relation based association rules mining
马春玲,李廉
摘要(Abstract)：
由于在实际的数据挖掘过程中容易出现无用的频集和冗余的规则 ,所以降低频集和规则的冗余度可大大提高挖掘的质量 ,这也是数据挖掘中一直被关注的问题 .提出了一个用等价类生成关联规则的方法 .算法主要在频集的基础上建立项集的等价关系 ,进而对项集划分等价类 ,同时将得到的关联规则划分为精确关联规则和近似关联规则两个集合 .通过等价类 ,不但可以很容易地生成所需要的关联规则 ,同时可以方便地判断数据之间依赖关系的强弱 .同时 ,项集的等价关系在实际应用中也很有利用价值 .算法最后的规则结果集剔除了由来自同一等价类中的项集的重复出现构成的冗余规则 ,从而得到了较小的关联规则集合
关键词(KeyWords)： 关联规则;频集;等价类;等价关系
基金项目(Foundation):
作者(Author): 马春玲,李廉
DOI: 10.13885/j.issn.0455-2059.2002.02.013
参考文献(References)：
- [1] AgrawalR,ImielinskiT,SwamiA N.Mining association rules between sets of items in large databases[A].PeterB,SushilJ.Proceedings of the1993ACM SIGMOD InternationalConference onManage-ment ofData[C].WashingtonD C:ACM Press,1993.207-216.
- [2] AgrawalR,SrikantR.Fast algorithms for mining association rules in large databases[A].JorgeB B,MatthiasJ,CarloZ.VL DB'94,Proceedings of the20 thInternationalConference onVeryL argeDataBases[C].Santiago deChile:MorganKaufmann,1994.487-499.
- [3] BrinS,MotwaniR,UllmanJD,et al.Dynamic itemset counting and implication rules for market bas-ket data[A].JoanP.ProceedingsACM SIGMOD InternationalConference onManagement ofData[C].Tucson,Arizona,USA:ACM Press,1997.265-264.
- [4] ParkJS,ChenM S,YuP S.An effective hash based algorithm for mining association rules[A].MichaelJC,DonovanA S.Proceedings of the1995ACM SIGMOD InternationalConference onManagement ofData[C].SanJose,California:ACM Press,1995.175-186.
- [5] HanJ,FuY.Discovery of multiple- level association rules from large databases[A].UmeshwarD,PeterM D G,ShojiroN.VL DB'95,Proceedings of21thInternationalConference onVeryL argeDataBases[C].Zurich,Switzerland:MorganKaufmann,1995.399-431.
- [6] SavasereA,OmiecinskiE,NavatheS.An efficient algorithm for mining association rules in largedatabases[A].UmeshwarD,PeterM D G,ShojiroN.VL DB'95,Proceedings of21thInternationaConference onVeryL argeDataBases[C].Zurich,Switzerland:MorganKaufmann,1995.432-443.
- [7] HanJ,PeiJ,YinJ.Mining frequentpatterns without candidate generation[A].WeidongC,TeffreyF N,PhilipA B.Proceeding2000ACM SIGMOD InternationalManagement ofData(SIGMOD'00)[C].Dal-las,TX:ACM Press,2000.1-12.
- [8] HanJ,FuY.Discovery of multiple- level association rules from large databases[A].UmeshwarD,PeterM D G,ShojiroN.VL DB'95,Proceedings of21thInternationalConference onVeryLargeDataBases[C].Zurich,Switzerland:MorganKaufmann,1995.420-431.
- [9] SrikantR,AgrawalR.Mining generalized association rules[A].UmeshwarD,PeterM D G,ShojiroN.VL DB'95,Proceedings of21thInternationalConference onVeryL argeDataBases[C].Zurich,Switzer-land:MorganKaufmann,1995.407-419.
- [10] PasquierN,BastideY,TaouilR,et al.Efficient mining of association rules using closed itemsets lat-tices[J].InformationSystems,1999,24(1):25-46