Presentation 2000/10/6
Matrix Clustering : a New Data Mining Algorithm for CRM
Shigeru Oyanagi, Kazuto Kubota, Akihiko Nakase,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A new data mining method named Matrix Clustering is proposed for CRM (Customer Relationship Managemnt). Matrix clustering is defined to generate a dense sub-matrix from a sparse binary matrix by exchanging rows and columns. We also propose a new fast algorithm named ping-pong algorithm for clustering large sparse matrix. The ping-pong algorithm is 1, 000 to 10, 000 times faster than a naive algorithm. This algorithm is applied to WWW access log analysis, and can find valuable relationshops between users and their accessed URL's..
Keyword(in Japanese) (See Japanese page)
Keyword(in English) data mining / clustering / CRM / ping-pong algorithm / Web access log analysis
Paper # DE2000-94
Date of Issue

Conference Information
Committee DE
Conference Date 2000/10/6(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Data Engineering (DE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Matrix Clustering : a New Data Mining Algorithm for CRM
Sub Title (in English)
Keyword(1) data mining
Keyword(2) clustering
Keyword(3) CRM
Keyword(4) ping-pong algorithm
Keyword(5) Web access log analysis
1st Author's Name Shigeru Oyanagi
1st Author's Affiliation RWCP TOSHIBA Parallel Application Laboratory Toshiba R&D Center()
2nd Author's Name Kazuto Kubota
2nd Author's Affiliation RWCP TOSHIBA Parallel Application Laboratory Toshiba R&D Center
3rd Author's Name Akihiko Nakase
3rd Author's Affiliation RWCP TOSHIBA Parallel Application Laboratory Toshiba R&D Center
Date 2000/10/6
Paper # DE2000-94
Volume (vol) vol.100
Number (no) 351
Page pp.pp.-
#Pages 8
Date of Issue