Presentation 2006-09-15
Topic analysis of natural language documents based on decision tree algorithm
Yusuke FURUHATA, Toshihiro NISHIZONO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper attempts to extract_specific document sets containing the same topic from masses of documents. The original documents are classified through a decision tree and the classification results are analyzed with the decision tree structure. Focusing on classification accuracy in each of decision tree leaves, the analysis yields several tens of specified nouns and pairs of nouns, of which existence identifies the class of each document. Then, topic identification ability of the nouns is ascertained using similarity and entropy of documents in each leaf. As a result, most of the nouns can specify document topics. The resulting topics are not sufficient for applying to new communication services. However, several parts of the decision tree can indicate efficient extraction of topic.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Decision Tree / Document Classification / Communication Services / Text Mining
Paper # CQ2006-54,OIS2006-41,IE2006-56
Date of Issue

Conference Information
Committee CQ
Conference Date 2006/9/8(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 Communication Quality (CQ)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Topic analysis of natural language documents based on decision tree algorithm
Sub Title (in English)
Keyword(1) Decision Tree
Keyword(2) Document Classification
Keyword(3) Communication Services
Keyword(4) Text Mining
1st Author's Name Yusuke FURUHATA
1st Author's Affiliation Graduate School of Engineering, Nihon University()
2nd Author's Name Toshihiro NISHIZONO
2nd Author's Affiliation Graduate School of Engineering, Nihon University
Date 2006-09-15
Paper # CQ2006-54,OIS2006-41,IE2006-56
Volume (vol) vol.106
Number (no) 240
Page pp.pp.-
#Pages 6
Date of Issue