Presentation 2003/7/24
Discovery of Web Communities from Positive and Negative Examples
Tsuyoshi MURATA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Several attempts have been made for Web structure mining whose goals are to discover or to rank related Webpages based on the graph structure of hyperlinks. Discovery of Web communities, groups of related Web pages, is importantfor assisting users' information retrieval from the Web. As is the case with human communities, Web communities are notuniform and are overlapped with one another. This causes topic drift of hyperlink-based algorithms such as HITS, and makesthe identification of the boundaries of Web communities difficult. This paper proposes a method for identifying Webcommunities from some positive and negative examples. Since the boundary of a Web community is hard to define only frompositive examples, negative examples are used for limiting its boundary from outer side. Experimental results show that ournew method is effective for finding Web communities' boundaries in some cases.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Web community / Web structure mining / discovery / hyperlink
Paper # AI2003-16
Date of Issue

Conference Information
Committee AI
Conference Date 2003/7/24(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 Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Discovery of Web Communities from Positive and Negative Examples
Sub Title (in English)
Keyword(1) Web community
Keyword(2) Web structure mining
Keyword(3) discovery
Keyword(4) hyperlink
1st Author's Name Tsuyoshi MURATA
1st Author's Affiliation National Institute of Informatics:Japan Science and Technology Corporation()
Date 2003/7/24
Paper # AI2003-16
Volume (vol) vol.103
Number (no) 243
Page pp.pp.-
#Pages 6
Date of Issue