Sorry, you need to enable JavaScript to visit this website.

Knowledge and Data Engineering

Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity

Tzu-Hsuan Yang
Tzu-Hsuan Tseng

In this paper, we study language models based on recurrent neural networks on three databases in two languages. We implement basic recurrent neural networks (RNN) and refined RNNs with long short-term memory (LSTM) cells. We use the corpora of Penn Tree Bank (PTB) and AMI in English, and the Academia Sinica Balanced Corpus (ASBC) in Chinese. On ASBC, we investigate wordbased and character-based language models. For characterbased language models, we look into the cases where the inter-word space is treated or not treated as a token.

40_RNN.pdf

PDF icon 40_RNN (234 downloads)

Paper Details

Authors:
Tzu-Hsuan Yang
Tzu-Hsuan Tseng
Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen
Submitted On:
21 November 2016 - 10:21am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

40_RNN

(234 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen, "Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1285. Accessed: Aug. 21, 2018.
@article{1285-16,
url = {http://sigport.org/1285},
author = {Tzu-Hsuan Yang; Tzu-Hsuan Tseng; Chia-Ping Chen },
publisher = {IEEE SigPort},
title = {Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity},
year = {2016} }
TY - EJOUR
T1 - Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity
AU - Tzu-Hsuan Yang; Tzu-Hsuan Tseng; Chia-Ping Chen
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1285
ER -
Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen. (2016). Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity. IEEE SigPort. http://sigport.org/1285
Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen, 2016. Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity. Available at: http://sigport.org/1285.
Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen. (2016). "Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity." Web.
1. Tzu-Hsuan Yang, Tzu-Hsuan Tseng, Chia-Ping Chen. Recurrent Neural Network-based Language Models with Variation in Net Topology, Language, and Granularity [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1285

Verifying the Long-range Dependency of RNN Language Models

Tzu-Hsuan Tseng
Tzu-Hsuan Yang

It has been argued that recurrent neural network language models are better in capturing long-range dependency than n-gram language models. In this paper, we attempt to verify this claim by investigating the prediction accuracy and the perplexity of these language models as a function of word position, i.e., the position of a word in a sentence. It is expected that as word position increases, the advantage of using recurrent neural network language models over n-gram language models will become more and more evident.

Paper Details

Authors:
Tzu-Hsuan Tseng
Tzu-Hsuan Yang
Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen
Submitted On:
21 November 2016 - 10:24am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

41_ngram_rnn

(216 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen, "Verifying the Long-range Dependency of RNN Language Models", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1284. Accessed: Aug. 21, 2018.
@article{1284-16,
url = {http://sigport.org/1284},
author = {Tzu-Hsuan Tseng; Tzu-Hsuan Yang; Chia-Ping Chen },
publisher = {IEEE SigPort},
title = {Verifying the Long-range Dependency of RNN Language Models},
year = {2016} }
TY - EJOUR
T1 - Verifying the Long-range Dependency of RNN Language Models
AU - Tzu-Hsuan Tseng; Tzu-Hsuan Yang; Chia-Ping Chen
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1284
ER -
Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen. (2016). Verifying the Long-range Dependency of RNN Language Models. IEEE SigPort. http://sigport.org/1284
Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen, 2016. Verifying the Long-range Dependency of RNN Language Models. Available at: http://sigport.org/1284.
Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen. (2016). "Verifying the Long-range Dependency of RNN Language Models." Web.
1. Tzu-Hsuan Tseng, Tzu-Hsuan Yang, Chia-Ping Chen. Verifying the Long-range Dependency of RNN Language Models [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1284

History Question Classification and Representation for Chinese Gaokao

Ke Yu

In this paper, we propose a question representation based on entity labeling and question classification for a automatic question answering system of Chinese Gaokao history question. A CRF model is used for the entity labeling and SVM/CNN/LSTM models are tested for question classification. Our experiments show that CRF model provides a high performance when used to label informative entities out while neural networks has a promising performance for the question classification task.

80.pdf

PDF icon 80.pdf (252 downloads)

Paper Details

Authors:
Ke Yu
Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng
Submitted On:
21 November 2016 - 9:27pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

80.pdf

(252 downloads)

Keywords

Subscribe

Send Author a Private Message
[1] Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng, "History Question Classification and Representation for Chinese Gaokao", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1283. Accessed: Aug. 21, 2018.
@article{1283-16,
url = {http://sigport.org/1283},
author = {Ke Yu; Qiuzhi Liu; Yuqing Zheng; Tiejun Zhao; Dequan Zheng },
publisher = {IEEE SigPort},
title = {History Question Classification and Representation for Chinese Gaokao},
year = {2016} }
TY - EJOUR
T1 - History Question Classification and Representation for Chinese Gaokao
AU - Ke Yu; Qiuzhi Liu; Yuqing Zheng; Tiejun Zhao; Dequan Zheng
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1283
ER -
Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng. (2016). History Question Classification and Representation for Chinese Gaokao. IEEE SigPort. http://sigport.org/1283
Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng, 2016. History Question Classification and Representation for Chinese Gaokao. Available at: http://sigport.org/1283.
Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng. (2016). "History Question Classification and Representation for Chinese Gaokao." Web.
1. Ke Yu, Qiuzhi Liu, Yuqing Zheng, Tiejun Zhao, Dequan Zheng. History Question Classification and Representation for Chinese Gaokao [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1283

The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation

Kazuhide Yamamoto

This paper presents initial English-Tigrinya statistical machine translation (SMT) research. Tigrinya is a highly inflected Semitic language spoken in Eritrea and Ethiopia. Translation involving morphologically complex languages is challenged by factors including data sparseness and source-target word alignment. We try to address these problems through morphological segmentation of Tigrinya words. After segmentation the difference in token count dropped significantly from 37.7% to 0.1%. The out-of-vocabulary rate was reduced by 46%.

IALP-tig.pdf

PDF icon IALP-tig.pdf (253 downloads)

Paper Details

Authors:
Kazuhide Yamamoto
Yemane Tedla and Kazuhide Yamamoto
Submitted On:
21 November 2016 - 8:31am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

IALP-tig.pdf

(253 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Yemane Tedla and Kazuhide Yamamoto, "The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1281. Accessed: Aug. 21, 2018.
@article{1281-16,
url = {http://sigport.org/1281},
author = {Yemane Tedla and Kazuhide Yamamoto },
publisher = {IEEE SigPort},
title = {The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation},
year = {2016} }
TY - EJOUR
T1 - The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation
AU - Yemane Tedla and Kazuhide Yamamoto
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1281
ER -
Yemane Tedla and Kazuhide Yamamoto. (2016). The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation. IEEE SigPort. http://sigport.org/1281
Yemane Tedla and Kazuhide Yamamoto, 2016. The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation. Available at: http://sigport.org/1281.
Yemane Tedla and Kazuhide Yamamoto. (2016). "The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation." Web.
1. Yemane Tedla and Kazuhide Yamamoto. The Effect of Shallow Segmentation for English-Tigrinya Statistical Machine Translation [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1281

Word Sense Implantation as Orthographical Conversion

Kazuhide Yamamoto

We present a word sense disambiguation (WSD) tool of Japanese Hiragana words. Unlike other WSD tasks which output something like “sense #3” as result, our WSD task rewrites the target word into a Kanji word, which is a different orthography. This means that the task is also a kind of orthographical normalization as well as WSD. In this paper we present the task, our method, and the performance.

IALP-wsd.pdf

PDF icon IALP-wsd.pdf (239 downloads)

Paper Details

Authors:
Kazuhide Yamamoto
Kazuhide Yamamoto and Yuki Mikami
Submitted On:
21 November 2016 - 8:27am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

IALP-wsd.pdf

(239 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Kazuhide Yamamoto and Yuki Mikami, "Word Sense Implantation as Orthographical Conversion", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1280. Accessed: Aug. 21, 2018.
@article{1280-16,
url = {http://sigport.org/1280},
author = {Kazuhide Yamamoto and Yuki Mikami },
publisher = {IEEE SigPort},
title = {Word Sense Implantation as Orthographical Conversion},
year = {2016} }
TY - EJOUR
T1 - Word Sense Implantation as Orthographical Conversion
AU - Kazuhide Yamamoto and Yuki Mikami
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1280
ER -
Kazuhide Yamamoto and Yuki Mikami. (2016). Word Sense Implantation as Orthographical Conversion. IEEE SigPort. http://sigport.org/1280
Kazuhide Yamamoto and Yuki Mikami, 2016. Word Sense Implantation as Orthographical Conversion. Available at: http://sigport.org/1280.
Kazuhide Yamamoto and Yuki Mikami. (2016). "Word Sense Implantation as Orthographical Conversion." Web.
1. Kazuhide Yamamoto and Yuki Mikami. Word Sense Implantation as Orthographical Conversion [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1280

Detecting Representative Web Articles Using Heterogeneous Graphs

RICHENG XUAN

Paper Details

Authors:
RICHENG XUAN
Submitted On:
21 November 2016 - 3:13am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

slide.11.21_upload.pptx

(237 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] , "Detecting Representative Web Articles Using Heterogeneous Graphs", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1279. Accessed: Aug. 21, 2018.
@article{1279-16,
url = {http://sigport.org/1279},
author = { },
publisher = {IEEE SigPort},
title = {Detecting Representative Web Articles Using Heterogeneous Graphs},
year = {2016} }
TY - EJOUR
T1 - Detecting Representative Web Articles Using Heterogeneous Graphs
AU -
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1279
ER -
. (2016). Detecting Representative Web Articles Using Heterogeneous Graphs. IEEE SigPort. http://sigport.org/1279
, 2016. Detecting Representative Web Articles Using Heterogeneous Graphs. Available at: http://sigport.org/1279.
. (2016). "Detecting Representative Web Articles Using Heterogeneous Graphs." Web.
1. . Detecting Representative Web Articles Using Heterogeneous Graphs [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1279

Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus

Kahaerjiang Abiderexiti

Uyghur is minority language in China, it is one of the official languages in Xinjiang Uyghur Autonomous Region of China. More than 10 million people use Uyghur in their daily life and even on the Internet. However, lack of Uyghur entity relation corpus constrains relation extraction applications in Uyghur. In this paper, we describe annotation schemes for creating annotated corpus for Uyghur named entity and Uyghur named entity relation.

Paper Details

Authors:
Kahaerjiang Abiderexiti
Submitted On:
20 November 2016 - 3:27am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus

(170 downloads)

Keywords

Subscribe

Send Author a Private Message
[1] , "Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1277. Accessed: Aug. 21, 2018.
@article{1277-16,
url = {http://sigport.org/1277},
author = { },
publisher = {IEEE SigPort},
title = {Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus},
year = {2016} }
TY - EJOUR
T1 - Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus
AU -
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1277
ER -
. (2016). Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus. IEEE SigPort. http://sigport.org/1277
, 2016. Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus. Available at: http://sigport.org/1277.
. (2016). "Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus." Web.
1. . Annotation Schemes for Constructing Uyghur Named Entity Relation Corpus [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1277

Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank

shuqin zhu

Paper Details

Authors:
shuqin zhu
Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song
Submitted On:
20 November 2016 - 2:21am
Short Link:
Type:
Event:

Document Files

Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank.pdf

(182 downloads)

Keywords

Subscribe

Send Author a Private Message
[1] Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song, "Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1276. Accessed: Aug. 21, 2018.
@article{1276-16,
url = {http://sigport.org/1276},
author = {Shuqin Zhu;Yinxia Zhang;Weiming Peng;Jihua Song },
publisher = {IEEE SigPort},
title = {Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank},
year = {2016} }
TY - EJOUR
T1 - Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank
AU - Shuqin Zhu;Yinxia Zhang;Weiming Peng;Jihua Song
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1276
ER -
Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song. (2016). Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank. IEEE SigPort. http://sigport.org/1276
Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song, 2016. Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank. Available at: http://sigport.org/1276.
Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song. (2016). "Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank." Web.
1. Shuqin Zhu,Yinxia Zhang,Weiming Peng,Jihua Song. Construction of the Basic Sentence-pattern Instance Database Based on the International Chinese Textbook Treebank [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1276

Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation

Kazuhide Yamamoto
Kanji Takahashi

We have investigated the effect of normalizing Japanese orthographical variants into a uniform orthography on statistical machine translation (SMT) between Japanese and English. In Japanese, 10% of words have reportedly more than one orthographical variants, which is a promising fact for improving translation quality when we normalize these orthographical variants.

Paper Details

Authors:
Kazuhide Yamamoto
Kanji Takahashi
Kazuhide Yamamoto, Kanji Takahashi
Submitted On:
21 November 2016 - 8:28pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

15-IALP2016.pdf

(227 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Kazuhide Yamamoto, Kanji Takahashi, "Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1273. Accessed: Aug. 21, 2018.
@article{1273-16,
url = {http://sigport.org/1273},
author = {Kazuhide Yamamoto; Kanji Takahashi },
publisher = {IEEE SigPort},
title = {Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation},
year = {2016} }
TY - EJOUR
T1 - Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation
AU - Kazuhide Yamamoto; Kanji Takahashi
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1273
ER -
Kazuhide Yamamoto, Kanji Takahashi. (2016). Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation. IEEE SigPort. http://sigport.org/1273
Kazuhide Yamamoto, Kanji Takahashi, 2016. Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation. Available at: http://sigport.org/1273.
Kazuhide Yamamoto, Kanji Takahashi. (2016). "Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation." Web.
1. Kazuhide Yamamoto, Kanji Takahashi. Japanese Orthographical Normalization Does Not Work for Statistical Machine Translation [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1273

Fundamental Tools and Resource are Available for Vietnamese Analysis

Kanji Takahashi
Kazuhide Yamamoto

This paper presents our work on developing Vietnamese fundamental tools and a resource for analysis. These tools are for word segmentation and part-of-speech tagging, diacritics restoration, and orthographical variants dictionary. All of them have been either not publicly available so far or not attaining sufficient performance. We have developed the tools and released the tools to the public, in both software packages and web tools. For development, we utilize state-of-the-art methods and achieved high accuracy.

Paper Details

Authors:
Kanji Takahashi
Kazuhide Yamamoto
Kanji Takahash, Kazuhide Yamamoto
Submitted On:
21 November 2016 - 4:25am
Short Link:
Type:
Event:
Paper Code:
Document Year:
Cite

Document Files

68-IALP2016.pdf

(220 downloads)

Keywords

Additional Categories

Subscribe

Send Author a Private Message
[1] Kanji Takahash, Kazuhide Yamamoto, "Fundamental Tools and Resource are Available for Vietnamese Analysis", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1271. Accessed: Aug. 21, 2018.
@article{1271-16,
url = {http://sigport.org/1271},
author = {Kanji Takahash; Kazuhide Yamamoto },
publisher = {IEEE SigPort},
title = {Fundamental Tools and Resource are Available for Vietnamese Analysis},
year = {2016} }
TY - EJOUR
T1 - Fundamental Tools and Resource are Available for Vietnamese Analysis
AU - Kanji Takahash; Kazuhide Yamamoto
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1271
ER -
Kanji Takahash, Kazuhide Yamamoto. (2016). Fundamental Tools and Resource are Available for Vietnamese Analysis. IEEE SigPort. http://sigport.org/1271
Kanji Takahash, Kazuhide Yamamoto, 2016. Fundamental Tools and Resource are Available for Vietnamese Analysis. Available at: http://sigport.org/1271.
Kanji Takahash, Kazuhide Yamamoto. (2016). "Fundamental Tools and Resource are Available for Vietnamese Analysis." Web.
1. Kanji Takahash, Kazuhide Yamamoto. Fundamental Tools and Resource are Available for Vietnamese Analysis [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1271

Pages