【作者】 王晨； 朱永泰； 汪卫； 施伯乐；

【Author】 WANG Chen,ZHU Yong-Tai,WANG Wei,and SHI Bai-Le (Department of Computing and Information Technology,Fudan University,Shanghai 200433)

【机构】 复旦大学计算机与信息技术系；

【摘要】 在过去的几年中,频繁项集和频繁序列的挖掘算法已经被用于解决各类应用问题,其中包括经典的市场篮和DNA序列.而目前,一系列新型数据挖掘应用迅速崛起,诸如化合物结构和计算机网络结构等的挖掘问题.图作为一种通用的数据结构,它可以表达数据间复杂的联系.因此,复杂的频繁结构模式的挖掘问题就可以转化成频繁子图的挖掘问题.提出了一个新的高效的频繁子图挖掘算法AdeSap.算法深度遍历搜索空间,构建邻接边投影库,递归挖掘频繁子图,并在挖掘过程中不产生侯选项.实验证明,AdeSap优于现有的其他图挖掘算法.更多还原

【Abstract】 Over these years,many algorithms of frequent itemsets and sequences discovery have been used to solve various problems with broad applications,including analyses of Market Basket,DNA sequences, and so on.As data mining techniques and applications,e.g.chemistry and computer network,have been developing and arising rapidly,it is urgent to study a more general structured patterns mining problem.As a general data structure,graph can model complicated relations among data.Within that model,the problem of finding frequent patterns becomes that of discovering subgraphs which occur frequently over the entire graphset.In this paper,new approaches for frequent subgraphs mining are investigated and an efficient algorithm named AdeSap(i.e.adjacency-projected subgraph pattern mining) is proposed which discovers frequent subgraphs without candidate generation.AdeSap builds projected database for each pattern discovered and recurs to find frequent subgraphs from it in depth-first search.Detailed experiments are conducted to evaluate the performance of AdeSap with synthetic datasets.The experimental results show that the method outperforms gSpan,the best performing frequent subgraphs mining algorithm reported so far.更多还原