Presentation 2004/11/27
Proposal of M2VSM and Its Comparison with Conventional VSM(Text Mining I)(Joint Workshop of Vietnamese Society of AI, SIGKBS-JSAI, ICS-IPSJ, and IEICE-SIGAI on Active Mining)
TORU ISHIBASHI, YASUFUMI TAKAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Information retrieval based on Vector Space Model (VSM) only employs typical indexing terms contained in documents. For that reason, when we apply it to a specific field such as medicine, it can crowd the documents in the vector space, which makes it difficult to retrieve and cluster them. In this paper, modified VSM based on meta keywords such as adjectives and adverbs, which is called M2VSM (Meta keyword-based Modified VSM), is proposed for separating the crowded documents using meta keywords as additional value of indexing terms. Experimental results by applying M2VSM to Medline (medical literature database) show that it can separate documents crowded in the vector space.
Keyword(in Japanese) (See Japanese page)
Keyword(in English)
Paper # AI2004-19
Date of Issue

Conference Information
Committee AI
Conference Date 2004/11/27(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 ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Proposal of M2VSM and Its Comparison with Conventional VSM(Text Mining I)(Joint Workshop of Vietnamese Society of AI, SIGKBS-JSAI, ICS-IPSJ, and IEICE-SIGAI on Active Mining)
Sub Title (in English)
Keyword(1)
1st Author's Name TORU ISHIBASHI
1st Author's Affiliation Tokyo Metropolitan Institue of Technology()
2nd Author's Name YASUFUMI TAKAMA
2nd Author's Affiliation Tokyo Metropolitan Institue of Technology
Date 2004/11/27
Paper # AI2004-19
Volume (vol) vol.104
Number (no) 485
Page pp.pp.-
#Pages 6
Date of Issue