Paper Abstract and Keywords |
Presentation |
2009-09-14 11:20
Automatic Summarization Based on Extracting Important Sentences
-- An Experiment and Analysis on Fluctuation of Deciding Whether or Not this Sentence Is Important -- Shiori Yamamoto, Kazuhiro Uenosono, Keigo Nakamura, Kiyomi Shirakawa, Hideji Enokizu, Seiichi Komiya (Shibaura Inst. of Tech.) KBSE2009-19 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
The authors aim at realizing automatic summarization of tales of old Japan such as Momotaro, Kachikachiyama, etc. from the viewpoint of important sentences. Akashi et al [1] tried to create a model of the process, which human understood
the text of above-mentioned tales, on the basis of the Van Dijk's model[3]. In the text summarization process, however, they detected the following fact: Some persons insisted that this sentence should be deleted for summarization, but other persons that the same sentence should be not deleted. This case was not so rare instance. Therefore, the authors carried out an experiment to investigate where does the above-mentioned case come from. This paper describes an experimental methodology and a result of the experiment. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Automatic Summarization / Extracting Important Sentences / Event Indexing Model / Breaking of Continuity / Quantification Method of the Second Type / Fluctuation of Deciding / / |
Reference Info. |
IEICE Tech. Rep., vol. 109, no. 196, KBSE2009-19, pp. 13-18, Sept. 2009. |
Paper # |
KBSE2009-19 |
Date of Issue |
2009-09-07 (KBSE) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
KBSE2009-19 |
|