A distant supervised relation extraction model with two denoising strategies

Zikai ZHOU, Yi CAI*, Jingyun XU, Jiayuan XIE, Qing LI, Haoran XIE

*Corresponding author for this work

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)

Abstract

Distant supervised relation extraction has been an effective way to find relational facts from text. However, distant supervised method inevitably accompanies with wrongly labeled sentences. Noisy sentences lead to poor performance of relation extraction models. Though existing piecewise convolutional neural network model with sentence-level attention (PCNN+ATT) is an effective way to reduce the effect of noisy sentences, it still has two limitations. On one hand, it adopts a PCNN module as sentence encoder, which only captures local contextual features of words and might lose important information. On the other hand, it neglects the fact that not all words contribute equally to the semantics of sentences. To address these two issues, we propose a hierarchical attention-based bidirectional GRU (HA-BiGRU) model. For the first limitation, our model utilizes a BiGRU module in place of PCNN, so as to extract global contextual information. For the second limitation, our model combines word-level and sentence-level attention mechanisms, which help get accurate sentence representations. To further alleviate the wrongly labeling problem, we first calculate the co-occurrence probabilities (CP) between the shortest dependency path (SDP) and the relation labels. Based on these co-occurrence probabilities, two denoising strategies are proposed to reduce noise interference respectively from aspect of filtering labeled data and integrating CP information into model. Experimental results on the corpus of Freebase and New York Times (Freebase+NYT) show that the HA-BiGRU model outperforms baseline models, and the two co-occurrence probabilities based denoising strategies can improve robustness of HA-BiGRU model.

Original languageEnglish
Title of host publication2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728119854
ISBN (Print)9781728119861
DOIs
Publication statusPublished - Jul 2019
Externally publishedYes
Event2019 International Joint Conference on Neural Networks, IJCNN 2019 - Budapest, Hungary
Duration: 14 Jul 201919 Jul 2019

Publication series

NameProceedings of the International Joint Conference on Neural Networks
Volume2019-July

Conference

Conference2019 International Joint Conference on Neural Networks, IJCNN 2019
CountryHungary
CityBudapest
Period14/07/1919/07/19

Fingerprint

Labeling
Labels
Semantics
Neural networks

Bibliographical note

This work was supported by the Fundamental Research Funds for the Central Universities, SCUT (No. 2017ZD048, D2182480), the Science and Technology Planning Project of Guangdong Province (No.2017B050506004), the Science and Technology Program of Guangzhou (No. 201704030076,201802010027). The research described in this paper has been supported by a collaborative research grant from the Hong Kong Research Grants Council (project no. C1031-18G).

Cite this

ZHOU, Z., CAI, Y., XU, J., XIE, J., LI, Q., & XIE, H. (2019). A distant supervised relation extraction model with two denoising strategies. In 2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings [8852378] (Proceedings of the International Joint Conference on Neural Networks; Vol. 2019-July). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IJCNN.2019.8852378
ZHOU, Zikai ; CAI, Yi ; XU, Jingyun ; XIE, Jiayuan ; LI, Qing ; XIE, Haoran. / A distant supervised relation extraction model with two denoising strategies. 2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. (Proceedings of the International Joint Conference on Neural Networks).
@inproceedings{270cd2ed992e4512a382d1b2333bf9f2,
title = "A distant supervised relation extraction model with two denoising strategies",
abstract = "Distant supervised relation extraction has been an effective way to find relational facts from text. However, distant supervised method inevitably accompanies with wrongly labeled sentences. Noisy sentences lead to poor performance of relation extraction models. Though existing piecewise convolutional neural network model with sentence-level attention (PCNN+ATT) is an effective way to reduce the effect of noisy sentences, it still has two limitations. On one hand, it adopts a PCNN module as sentence encoder, which only captures local contextual features of words and might lose important information. On the other hand, it neglects the fact that not all words contribute equally to the semantics of sentences. To address these two issues, we propose a hierarchical attention-based bidirectional GRU (HA-BiGRU) model. For the first limitation, our model utilizes a BiGRU module in place of PCNN, so as to extract global contextual information. For the second limitation, our model combines word-level and sentence-level attention mechanisms, which help get accurate sentence representations. To further alleviate the wrongly labeling problem, we first calculate the co-occurrence probabilities (CP) between the shortest dependency path (SDP) and the relation labels. Based on these co-occurrence probabilities, two denoising strategies are proposed to reduce noise interference respectively from aspect of filtering labeled data and integrating CP information into model. Experimental results on the corpus of Freebase and New York Times (Freebase+NYT) show that the HA-BiGRU model outperforms baseline models, and the two co-occurrence probabilities based denoising strategies can improve robustness of HA-BiGRU model.",
author = "Zikai ZHOU and Yi CAI and Jingyun XU and Jiayuan XIE and Qing LI and Haoran XIE",
note = "This work was supported by the Fundamental Research Funds for the Central Universities, SCUT (No. 2017ZD048, D2182480), the Science and Technology Planning Project of Guangdong Province (No.2017B050506004), the Science and Technology Program of Guangzhou (No. 201704030076,201802010027). The research described in this paper has been supported by a collaborative research grant from the Hong Kong Research Grants Council (project no. C1031-18G).",
year = "2019",
month = "7",
doi = "10.1109/IJCNN.2019.8852378",
language = "English",
isbn = "9781728119861",
series = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings",
address = "United States",

}

ZHOU, Z, CAI, Y, XU, J, XIE, J, LI, Q & XIE, H 2019, A distant supervised relation extraction model with two denoising strategies. in 2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings., 8852378, Proceedings of the International Joint Conference on Neural Networks, vol. 2019-July, Institute of Electrical and Electronics Engineers Inc., 2019 International Joint Conference on Neural Networks, IJCNN 2019, Budapest, Hungary, 14/07/19. https://doi.org/10.1109/IJCNN.2019.8852378

A distant supervised relation extraction model with two denoising strategies. / ZHOU, Zikai; CAI, Yi; XU, Jingyun; XIE, Jiayuan; LI, Qing; XIE, Haoran.

2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. 8852378 (Proceedings of the International Joint Conference on Neural Networks; Vol. 2019-July).

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)

TY - GEN

T1 - A distant supervised relation extraction model with two denoising strategies

AU - ZHOU, Zikai

AU - CAI, Yi

AU - XU, Jingyun

AU - XIE, Jiayuan

AU - LI, Qing

AU - XIE, Haoran

N1 - This work was supported by the Fundamental Research Funds for the Central Universities, SCUT (No. 2017ZD048, D2182480), the Science and Technology Planning Project of Guangdong Province (No.2017B050506004), the Science and Technology Program of Guangzhou (No. 201704030076,201802010027). The research described in this paper has been supported by a collaborative research grant from the Hong Kong Research Grants Council (project no. C1031-18G).

PY - 2019/7

Y1 - 2019/7

N2 - Distant supervised relation extraction has been an effective way to find relational facts from text. However, distant supervised method inevitably accompanies with wrongly labeled sentences. Noisy sentences lead to poor performance of relation extraction models. Though existing piecewise convolutional neural network model with sentence-level attention (PCNN+ATT) is an effective way to reduce the effect of noisy sentences, it still has two limitations. On one hand, it adopts a PCNN module as sentence encoder, which only captures local contextual features of words and might lose important information. On the other hand, it neglects the fact that not all words contribute equally to the semantics of sentences. To address these two issues, we propose a hierarchical attention-based bidirectional GRU (HA-BiGRU) model. For the first limitation, our model utilizes a BiGRU module in place of PCNN, so as to extract global contextual information. For the second limitation, our model combines word-level and sentence-level attention mechanisms, which help get accurate sentence representations. To further alleviate the wrongly labeling problem, we first calculate the co-occurrence probabilities (CP) between the shortest dependency path (SDP) and the relation labels. Based on these co-occurrence probabilities, two denoising strategies are proposed to reduce noise interference respectively from aspect of filtering labeled data and integrating CP information into model. Experimental results on the corpus of Freebase and New York Times (Freebase+NYT) show that the HA-BiGRU model outperforms baseline models, and the two co-occurrence probabilities based denoising strategies can improve robustness of HA-BiGRU model.

AB - Distant supervised relation extraction has been an effective way to find relational facts from text. However, distant supervised method inevitably accompanies with wrongly labeled sentences. Noisy sentences lead to poor performance of relation extraction models. Though existing piecewise convolutional neural network model with sentence-level attention (PCNN+ATT) is an effective way to reduce the effect of noisy sentences, it still has two limitations. On one hand, it adopts a PCNN module as sentence encoder, which only captures local contextual features of words and might lose important information. On the other hand, it neglects the fact that not all words contribute equally to the semantics of sentences. To address these two issues, we propose a hierarchical attention-based bidirectional GRU (HA-BiGRU) model. For the first limitation, our model utilizes a BiGRU module in place of PCNN, so as to extract global contextual information. For the second limitation, our model combines word-level and sentence-level attention mechanisms, which help get accurate sentence representations. To further alleviate the wrongly labeling problem, we first calculate the co-occurrence probabilities (CP) between the shortest dependency path (SDP) and the relation labels. Based on these co-occurrence probabilities, two denoising strategies are proposed to reduce noise interference respectively from aspect of filtering labeled data and integrating CP information into model. Experimental results on the corpus of Freebase and New York Times (Freebase+NYT) show that the HA-BiGRU model outperforms baseline models, and the two co-occurrence probabilities based denoising strategies can improve robustness of HA-BiGRU model.

UR - http://www.scopus.com/inward/record.url?scp=85073193270&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2019.8852378

DO - 10.1109/IJCNN.2019.8852378

M3 - Conference paper (refereed)

AN - SCOPUS:85073193270

SN - 9781728119861

T3 - Proceedings of the International Joint Conference on Neural Networks

BT - 2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

ZHOU Z, CAI Y, XU J, XIE J, LI Q, XIE H. A distant supervised relation extraction model with two denoising strategies. In 2019 International Joint Conference on Neural Networks, IJCNN 2019 : proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. 8852378. (Proceedings of the International Joint Conference on Neural Networks). https://doi.org/10.1109/IJCNN.2019.8852378