IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2014-11-17 17:00
[Poster Presentation] Reconstructive Neural Net based on Context Free Grammar
Takuo Hamaguchi, Masashi Shimbo, Yuji Matsumoto (NAIST) IBISML2014-48
Abstract (in Japanese) (See Japanese page) 
(in English) There are two aspects about Neural Language Model,one aspect, `Encode', obtains a vector from a sequence and the other aspect, `Decode', obtains a sequence from a vector.Recurrent Neural Network (Recurrent NN) is a kind of `Decode' model which repeats prediction and transition from vector-represented state through time.However, a tree structure based model such as Context Free Grammar is more natural for natural languate tha a time-series structure based model.This paper proposes new tree structure `Decode' Model called Reconstrucitve Neural Network (Reconstructive NN).We conduct the following four experiemtns to show this model's ability:comparision of Recurrent NN and Reconstructive NN,relationship between the accuracy and the Distribution of word length included training data,relationship between the vector dimension and the words length,and the ability of Reconstructive NN comes from natural language.
Keyword (in Japanese) (See Japanese page) 
(in English) Deep Learning / Distributed Representation / Recurrent Neural Network / Recursive Nerural Network / / / /  
Reference Info. IEICE Tech. Rep., vol. 114, no. 306, IBISML2014-48, pp. 103-110, Nov. 2014.
Paper # IBISML2014-48 
Date of Issue 2014-11-10 (IBISML) 
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 IBISML2014-48

Conference Information
Committee IBISML  
Conference Date 2014-11-17 - 2014-11-19 
Place (in Japanese) (See Japanese page) 
Place (in English) Nagoya Univ. 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To IBISML 
Conference Code 2014-11-IBISML 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Reconstructive Neural Net based on Context Free Grammar 
Sub Title (in English)  
Keyword(1) Deep Learning  
Keyword(2) Distributed Representation  
Keyword(3) Recurrent Neural Network  
Keyword(4) Recursive Nerural Network  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Takuo Hamaguchi  
1st Author's Affiliation Nara Institute of Science and Technology (NAIST)
2nd Author's Name Masashi Shimbo  
2nd Author's Affiliation Nara Institute of Science and Technology (NAIST)
3rd Author's Name Yuji Matsumoto  
3rd Author's Affiliation Nara Institute of Science and Technology (NAIST)
4th Author's Name  
4th Author's Affiliation ()
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker Author-1 
Date Time 2014-11-17 17:00:00 
Presentation Time 180 minutes 
Registration for IBISML 
Paper # IBISML2014-48 
Volume (vol) vol.114 
Number (no) no.306 
Page pp.103-110 
#Pages
Date of Issue 2014-11-10 (IBISML) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan