Task-oriented domain-specific meta-embedding for text classification

Xin Wu; Yi Cai; Qing Li; Tao Wang; Kai Yang

doi:10.18653/v1/2020.emnlp-main.282

Task-oriented domain-specific meta-embedding for text classification

Xin Wu, Yi Cai^*, Qing Li, Tao Wang, Kai Yang

^*Corresponding author for this work

Department of Information Systems

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

6 Citations (Scopus)

Abstract

Meta-embedding learning, which combines complementary information in different word embeddings, have shown superior performances across different Natural Language Processing tasks. However, domain-specific knowledge is still ignored by existing meta-embedding methods, which results in unstable performances across specific domains. Moreover, the importance of general and domain word embeddings is related to downstream tasks, how to regularize meta-embedding to adapt downstream tasks is an unsolved problem. In this paper, we propose a method to incorporate both domain-specific and task-oriented information into meta-embeddings. We conducted extensive experiments on four text classification datasets and the results show the effectiveness of our proposed method.

Original language	English
Title of host publication	Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing
Editors	Bonnie Webber, Trevor Cohn, Yulan He, Yang Li
Publisher	Association for Computational Linguistics
Pages	3508-3513
ISBN (Print)	9781952148606
DOIs	https://doi.org/10.18653/v1/2020.emnlp-main.282
Publication status	Published - Nov 2020
Event	2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020) - Virtual Duration: 16 Nov 2020 → 20 Nov 2020 https://2020.emnlp.org/

Publication series

Name	EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference

Conference

Conference	2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020)
Period	16/11/20 → 20/11/20
Internet address	https://2020.emnlp.org/

Bibliographical note

Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).

Access Output

10.18653/v1/2020.emnlp-main.282

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Wu, X., Cai, Y., Li, Q., Wang, T., & Yang, K. (2020). Task-oriented domain-specific meta-embedding for text classification. In B. Webber, T. Cohn, Y. He, & Y. Li (Eds.), Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (pp. 3508-3513). (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference). Association for Computational Linguistics. https://doi.org/10.18653/v1/2020.emnlp-main.282

Wu, Xin ; Cai, Yi ; Li, Qing et al. / Task-oriented domain-specific meta-embedding for text classification. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. editor / Bonnie Webber ; Trevor Cohn ; Yulan He ; Yang Li. Association for Computational Linguistics, 2020. pp. 3508-3513 (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference).

@inproceedings{cfd7590758494ddbaf31b500a81b07a4,

title = "Task-oriented domain-specific meta-embedding for text classification",

abstract = "Meta-embedding learning, which combines complementary information in different word embeddings, have shown superior performances across different Natural Language Processing tasks. However, domain-specific knowledge is still ignored by existing meta-embedding methods, which results in unstable performances across specific domains. Moreover, the importance of general and domain word embeddings is related to downstream tasks, how to regularize meta-embedding to adapt downstream tasks is an unsolved problem. In this paper, we propose a method to incorporate both domain-specific and task-oriented information into meta-embeddings. We conducted extensive experiments on four text classification datasets and the results show the effectiveness of our proposed method.",

author = "Xin Wu and Yi Cai and Qing Li and Tao Wang and Kai Yang",

note = "Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).; 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020) ; Conference date: 16-11-2020 Through 20-11-2020",

year = "2020",

month = nov,

doi = "10.18653/v1/2020.emnlp-main.282",

language = "English",

isbn = "9781952148606",

series = "EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference",

publisher = "Association for Computational Linguistics",

pages = "3508--3513",

editor = "Bonnie Webber and Trevor Cohn and Yulan He and Yang Li",

booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing",

address = "United States",

url = "https://2020.emnlp.org/",

}

Wu, X, Cai, Y, Li, Q, Wang, T & Yang, K 2020, Task-oriented domain-specific meta-embedding for text classification. in B Webber, T Cohn, Y He & Y Li (eds), Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference, Association for Computational Linguistics, pp. 3508-3513, 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020), 16/11/20. https://doi.org/10.18653/v1/2020.emnlp-main.282

Task-oriented domain-specific meta-embedding for text classification. / Wu, Xin; Cai, Yi; Li, Qing et al.
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. ed. / Bonnie Webber; Trevor Cohn; Yulan He; Yang Li. Association for Computational Linguistics, 2020. p. 3508-3513 (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Task-oriented domain-specific meta-embedding for text classification

AU - Wu, Xin

AU - Cai, Yi

AU - Li, Qing

AU - Wang, Tao

AU - Yang, Kai

N1 - Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).

PY - 2020/11

Y1 - 2020/11

N2 - Meta-embedding learning, which combines complementary information in different word embeddings, have shown superior performances across different Natural Language Processing tasks. However, domain-specific knowledge is still ignored by existing meta-embedding methods, which results in unstable performances across specific domains. Moreover, the importance of general and domain word embeddings is related to downstream tasks, how to regularize meta-embedding to adapt downstream tasks is an unsolved problem. In this paper, we propose a method to incorporate both domain-specific and task-oriented information into meta-embeddings. We conducted extensive experiments on four text classification datasets and the results show the effectiveness of our proposed method.

AB - Meta-embedding learning, which combines complementary information in different word embeddings, have shown superior performances across different Natural Language Processing tasks. However, domain-specific knowledge is still ignored by existing meta-embedding methods, which results in unstable performances across specific domains. Moreover, the importance of general and domain word embeddings is related to downstream tasks, how to regularize meta-embedding to adapt downstream tasks is an unsolved problem. In this paper, we propose a method to incorporate both domain-specific and task-oriented information into meta-embeddings. We conducted extensive experiments on four text classification datasets and the results show the effectiveness of our proposed method.

UR - http://www.scopus.com/inward/record.url?scp=85118435222&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85118435222&origin=recordpage

U2 - 10.18653/v1/2020.emnlp-main.282

DO - 10.18653/v1/2020.emnlp-main.282

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 9781952148606

T3 - EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference

SP - 3508

EP - 3513

BT - Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing

A2 - Webber, Bonnie

A2 - Cohn, Trevor

A2 - He, Yulan

A2 - Li, Yang

PB - Association for Computational Linguistics

T2 - 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020)

Y2 - 16 November 2020 through 20 November 2020

ER -

Wu X, Cai Y, Li Q, Wang T, Yang K. Task-oriented domain-specific meta-embedding for text classification. In Webber B, Cohn T, He Y, Li Y, editors, Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. 2020. p. 3508-3513. (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference). doi: 10.18653/v1/2020.emnlp-main.282

Task-oriented domain-specific meta-embedding for text classification

Abstract

Publication series

Conference

Bibliographical note

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this