Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task

Yuming Li; Pin Ni; Gangmin Li; Victor Chang

Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task

Yuming Li, Pin Ni, Gangmin Li, Victor Chang

Centre for Digital Innovation

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Citations (Scopus)

74 Downloads (Pure)

Abstract

Relation Extraction is an important sub-task in the field of information extraction. Its goal is to identify entities from text and extract semantic relationships between entities. However, the current Relationship Extraction task based on deep learning methods generally have practical problems such as insufficient amount of manually labeled data, so training under weak supervision has become a big challenge. Distant Supervision is a novel idea that can automatically annotate a large number of unlabeled data based on a small amount of labeled data. Based on this idea, this paper proposes a method combining the Piecewise Convolutional Neural Networks and Attention mechanism for automatically annotating the data of Relation Extraction task. The experiments proved that the proposed method achieved the highest precision is 76.24% on NYT-FB (New York Times-Freebase) dataset (top 100 relation categories). The results show that the proposed method performed better than CNN-based models in most cases.

Original language	English
Title of host publication	COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk
Editors	Reinhold Behringer, Victor Chang
Publisher	SciTePress
Pages	53-62
Number of pages	10
ISBN (Electronic)	9789897584275
Publication status	Published - 8 May 2020
Event	5th International Conference on Complexity, Future Information Systems and Risk - Virtual, Online Duration: 8 May 2020 → 9 May 2020

Publication series

Name	COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk

Conference

Conference	5th International Conference on Complexity, Future Information Systems and Risk
Abbreviated title	COMPLEXIS 2020
City	Virtual, Online
Period	8/05/20 → 9/05/20

Bibliographical note

Funding Information:
We are grateful to VC Research (Funding No. VCR 0000040) to support this work.

Publisher Copyright:
© 2020 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

Access to Document

Full textAccepted author manuscript, 352 KBLicence: CC BY-NC-ND

https://www.scitepress.org/Papers/2020/95827/95827.pdf

Cite this

Li, Y., Ni, P., Li, G., & Chang, V. (2020). Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task. In R. Behringer, & V. Chang (Eds.), COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk (pp. 53-62). (COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk). SciTePress. https://www.scitepress.org/Papers/2020/95827/95827.pdf

Li, Yuming ; Ni, Pin ; Li, Gangmin et al. / Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task. COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk. editor / Reinhold Behringer ; Victor Chang. SciTePress, 2020. pp. 53-62 (COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk).

@inproceedings{3aa7199136ea439f89ade348d987f5fe,

title = "Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task",

abstract = "Relation Extraction is an important sub-task in the field of information extraction. Its goal is to identify entities from text and extract semantic relationships between entities. However, the current Relationship Extraction task based on deep learning methods generally have practical problems such as insufficient amount of manually labeled data, so training under weak supervision has become a big challenge. Distant Supervision is a novel idea that can automatically annotate a large number of unlabeled data based on a small amount of labeled data. Based on this idea, this paper proposes a method combining the Piecewise Convolutional Neural Networks and Attention mechanism for automatically annotating the data of Relation Extraction task. The experiments proved that the proposed method achieved the highest precision is 76.24% on NYT-FB (New York Times-Freebase) dataset (top 100 relation categories). The results show that the proposed method performed better than CNN-based models in most cases.",

author = "Yuming Li and Pin Ni and Gangmin Li and Victor Chang",

note = "Funding Information: We are grateful to VC Research (Funding No. VCR 0000040) to support this work. Publisher Copyright: {\textcopyright} 2020 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved; 5th International Conference on Complexity, Future Information Systems and Risk, COMPLEXIS 2020 ; Conference date: 08-05-2020 Through 09-05-2020",

year = "2020",

month = may,

day = "8",

language = "English",

series = "COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk",

publisher = "SciTePress",

pages = "53--62",

editor = "Reinhold Behringer and Victor Chang",

booktitle = "COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk",

}

Li, Y, Ni, P, Li, G & Chang, V 2020, Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task. in R Behringer & V Chang (eds), COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk. COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk, SciTePress, pp. 53-62, 5th International Conference on Complexity, Future Information Systems and Risk, Virtual, Online, 8/05/20. <https://www.scitepress.org/Papers/2020/95827/95827.pdf>

Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task. / Li, Yuming; Ni, Pin; Li, Gangmin et al.
COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk. ed. / Reinhold Behringer; Victor Chang. SciTePress, 2020. p. 53-62 (COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task

AU - Li, Yuming

AU - Ni, Pin

AU - Li, Gangmin

AU - Chang, Victor

N1 - Funding Information: We are grateful to VC Research (Funding No. VCR 0000040) to support this work. Publisher Copyright: © 2020 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

PY - 2020/5/8

Y1 - 2020/5/8

N2 - Relation Extraction is an important sub-task in the field of information extraction. Its goal is to identify entities from text and extract semantic relationships between entities. However, the current Relationship Extraction task based on deep learning methods generally have practical problems such as insufficient amount of manually labeled data, so training under weak supervision has become a big challenge. Distant Supervision is a novel idea that can automatically annotate a large number of unlabeled data based on a small amount of labeled data. Based on this idea, this paper proposes a method combining the Piecewise Convolutional Neural Networks and Attention mechanism for automatically annotating the data of Relation Extraction task. The experiments proved that the proposed method achieved the highest precision is 76.24% on NYT-FB (New York Times-Freebase) dataset (top 100 relation categories). The results show that the proposed method performed better than CNN-based models in most cases.

AB - Relation Extraction is an important sub-task in the field of information extraction. Its goal is to identify entities from text and extract semantic relationships between entities. However, the current Relationship Extraction task based on deep learning methods generally have practical problems such as insufficient amount of manually labeled data, so training under weak supervision has become a big challenge. Distant Supervision is a novel idea that can automatically annotate a large number of unlabeled data based on a small amount of labeled data. Based on this idea, this paper proposes a method combining the Piecewise Convolutional Neural Networks and Attention mechanism for automatically annotating the data of Relation Extraction task. The experiments proved that the proposed method achieved the highest precision is 76.24% on NYT-FB (New York Times-Freebase) dataset (top 100 relation categories). The results show that the proposed method performed better than CNN-based models in most cases.

UR - http://www.scopus.com/inward/record.url?scp=85090380097&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85090380097

T3 - COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk

SP - 53

EP - 62

BT - COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk

A2 - Behringer, Reinhold

A2 - Chang, Victor

PB - SciTePress

T2 - 5th International Conference on Complexity, Future Information Systems and Risk

Y2 - 8 May 2020 through 9 May 2020

ER -

Li Y, Ni P, Li G, Chang V. Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task. In Behringer R, Chang V, editors, COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk. SciTePress. 2020. p. 53-62. (COMPLEXIS 2020 - Proceedings of the 5th International Conference on Complexity, Future Information Systems and Risk).

Effective Piecewise CNN with attention mechanism for distant supervision on relation extraction task

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this