Presentation 2004/2/12
Question Answering System for Document Images : IQAS (Thought and Language)
Shota FUKUSHIMA, Kohichi KISE, Keinosuke MATSUMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, we propose a question answering system for document images. Question answering systems are to extract from document databases exact parts of articles that directly answer the questions given by a user. Most of conventional question answering systems deal with electronic texts. However a great number of document images exist in electronic library. Such document images have unique information which is not obtained from electronic texts. In order to utilize these document images, we propose a question answering system for document images : IQAS (document Image Question Answering System). A characteristic point of IQAS is that it returns exact parts of document images based on 2D position information of terms. The experimental results on 1,253 pages (197 documents) with 20 queries show that MRR (Mean Reciprocal Rank) is 0.45.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) quesion answering system / document image / 2D denstiy distributions of terms / layout information
Paper # TL2003-31,PRMU2003-217
Date of Issue

Conference Information
Committee TL
Conference Date 2004/2/12(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 Thought and Language (TL)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Question Answering System for Document Images : IQAS (Thought and Language)
Sub Title (in English)
Keyword(1) quesion answering system
Keyword(2) document image
Keyword(3) 2D denstiy distributions of terms
Keyword(4) layout information
1st Author's Name Shota FUKUSHIMA
1st Author's Affiliation Dept. of Computer and Systems Sciences, Osaka Prefecture University()
2nd Author's Name Kohichi KISE
2nd Author's Affiliation Dept. of Computer and Systems Sciences, Graduate School of ENG., Osaka Prefecture University
3rd Author's Name Keinosuke MATSUMOTO
3rd Author's Affiliation Dept. of Computer and Systems Sciences, Graduate School of ENG., Osaka Prefecture University
Date 2004/2/12
Paper # TL2003-31,PRMU2003-217
Volume (vol) vol.103
Number (no) 656
Page pp.pp.-
#Pages 6
Date of Issue