Presentation 2011-07-08
Extracting evaluative tuples with structure learning
Hajime MORITA, Hiroya TAKAMURA, Manabu OKUMURA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we intend to extract tuples of evaluative expression on targets from web texts. An evaluative tuple contains three types of elements, target, attribute and evaluation. We expand the scope of elements to continuous words instead of a single word to collect more diverse types of expressions. We propose a new method to learn to extract tuples taking into account dependency within expressions, because the target and attribute have a high correlation with evaluation each other. We will discuss an issue and resolution based on the results of experiments on the tagged web texts corpus.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Evaluative expression extraction / Evaluative tuple extraction / Structured output learning / online learning
Paper # NLC2011-17
Date of Issue

Conference Information
Committee NLC
Conference Date 2011/6/30(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 Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Extracting evaluative tuples with structure learning
Sub Title (in English)
Keyword(1) Evaluative expression extraction
Keyword(2) Evaluative tuple extraction
Keyword(3) Structured output learning
Keyword(4) online learning
1st Author's Name Hajime MORITA
1st Author's Affiliation Tokyo Institute of Technology()
2nd Author's Name Hiroya TAKAMURA
2nd Author's Affiliation Precision and Intelligence Laboratory, Tokyo Institute of Technology
3rd Author's Name Manabu OKUMURA
3rd Author's Affiliation Precision and Intelligence Laboratory, Tokyo Institute of Technology
Date 2011-07-08
Paper # NLC2011-17
Volume (vol) vol.111
Number (no) 119
Page pp.pp.-
#Pages 5
Date of Issue