SWSPMiner:Efficient Mining of Weighted Sequential Patterns from Traversals on Weighted Directed Graph Using Statistical Theory
To solve the problem of mining weighted traversal patterns (WTPs) with noisy weight from weighted directed graph (WDG),an effective algorithm,called SWSPMiner (Statistical theory-based Weighted Sequential Patterns Miner),is proposed.The algorithm undergoes two phases to discover WSPs from the traversals on WDG.In the first phase,it adopts the weights confidence interval (C1) to delete the vertices with noisy weights from the traversal database (TDB),which reduce remarkably the size of TDB.In the second phase,based on the property that the items in a traversal pattern are consecutive,the algorithm regards a traversal pattern as a sequence pattern.Then the algorithm adopts an improved weighted prefix-projected pattern growth approach to decompose the task of mining original sequence database into a series of smaller tasks of mining locally projected database and pushes the weight constraint into the mining process so as to efficiently discover fewer but more important WTPs.Comprehensive experimental results show that the algorithm is efficient and scalable for mining sequential patterns from traversals on the WDG.Moreover,the algorithm can be applied to various applications which can be modeled as a WDG.
Statistical Theory Weighted Directed Graph Traversal Pattern Mining Sequential Pattern Mining
Runian Geng Wenbo Xu Xiangjun Dong
School of Information Technology,Jiangnan University;Wuxi,Jiangsu 214122,China School of Information School of Information Technology,Jiangnan University;Wuxi,Jiangsu 214122,China School of Information Science and Technology,Shandong Institute of Light Industry Jinan,Shandong 250
国际会议
大连
英文
473-481
2008-07-27（万方平台首次上网日期，不代表论文的发表时间）