Leveraging linguistic resources for improving neural text classification

Ming Liu; Gholamreza Haffari; Wray Buntine; Michelle Ananda-Rajah

Leveraging linguistic resources for improving neural text classification

Ming Liu, Gholamreza Haffari, Wray Buntine, Michelle Ananda-Rajah

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

Abstract

This paper presents a deep linguistic attentional framework which incorporates word level concept information into neural classification models. While learning neural classification models often requires a large amount of labelled data, linguistic concept information can be obtained from external knowledge, such as pre-trained word embeddings, WordNet for common text and MetaMap for biomedical text. We explore two different ways of incorporating word level concept annotations, and show that leveraging concept annotation scan boost the model performance and reduce the need for large amounts of labelled data. Experiments on various data sets validate the effectiveness of the proposed method.

Original language	English
Title of host publication	Australasian Language Technology Association Workshop, ALTA 2017
Subtitle of host publication	6–8 December 2017, Brisbane, Australia, Proceedings
Editors	Jojo Sze-Meng Wong, Gholamreza Haffari
Place of Publication	Melbourne Vic Australia
Publisher	Australian Language Technology Association (ALTA)
Pages	34-42
Number of pages	9
Volume	15
Edition	2017
Publication status	Published - 2017
Event	Australasian Language Technology Association Workshop 2017 - Queensland University of Technology, Brisbane, Australia Duration: 6 Dec 2017 → 8 Dec 2017 Conference number: 15th https://www.aclweb.org/anthology/events/alta-2017/ (Proceedings)

Conference

Conference	Australasian Language Technology Association Workshop 2017
Abbreviated title	ALTAW 2017
Country/Territory	Australia
City	Brisbane
Period	6/12/17 → 8/12/17
Other	In 2017, the Australasian Language Technology Association Workshop (ALTA) will be held at the Queensland University of Technology in Brisbane. The workshop is the key forum in Australia and New Zealand for sharing research results in natural language processing and computational linguistics, with featured keynote speakers, as well as presentations and posters from student, industry, and early-career researchers. ALTA 2017 is co-located with ADCS 2017.
Internet address	https://www.aclweb.org/anthology/events/alta-2017/ (Proceedings)

Cite this

Liu, M., Haffari, G., Buntine, W., & Ananda-Rajah, M. (2017). Leveraging linguistic resources for improving neural text classification. In J. S.-M. Wong, & G. Haffari (Eds.), Australasian Language Technology Association Workshop, ALTA 2017: 6–8 December 2017, Brisbane, Australia, Proceedings (2017 ed., Vol. 15, pp. 34-42). Australian Language Technology Association (ALTA).

Liu, Ming ; Haffari, Gholamreza ; Buntine, Wray et al. / Leveraging linguistic resources for improving neural text classification. Australasian Language Technology Association Workshop, ALTA 2017: 6–8 December 2017, Brisbane, Australia, Proceedings . editor / Jojo Sze-Meng Wong ; Gholamreza Haffari. Vol. 15 2017. ed. Melbourne Vic Australia : Australian Language Technology Association (ALTA), 2017. pp. 34-42

@inproceedings{3a8ca734c3484736ae94cd250d70a557,

title = "Leveraging linguistic resources for improving neural text classification",

abstract = "This paper presents a deep linguistic attentional framework which incorporates word level concept information into neural classification models. While learning neural classification models often requires a large amount of labelled data, linguistic concept information can be obtained from external knowledge, such as pre-trained word embeddings, WordNet for common text and MetaMap for biomedical text. We explore two different ways of incorporating word level concept annotations, and show that leveraging concept annotation scan boost the model performance and reduce the need for large amounts of labelled data. Experiments on various data sets validate the effectiveness of the proposed method.",

author = "Ming Liu and Gholamreza Haffari and Wray Buntine and Michelle Ananda-Rajah",

year = "2017",

language = "English",

volume = "15",

pages = "34--42",

editor = "Wong, {Jojo Sze-Meng} and Gholamreza Haffari",

booktitle = "Australasian Language Technology Association Workshop, ALTA 2017",

publisher = "Australian Language Technology Association (ALTA)",

edition = "2017",

note = "Australasian Language Technology Association Workshop 2017, ALTAW 2017 ; Conference date: 06-12-2017 Through 08-12-2017",

url = "https://www.aclweb.org/anthology/events/alta-2017/",

}

Liu, M, Haffari, G, Buntine, W & Ananda-Rajah, M 2017, Leveraging linguistic resources for improving neural text classification. in JS-M Wong & G Haffari (eds), Australasian Language Technology Association Workshop, ALTA 2017: 6–8 December 2017, Brisbane, Australia, Proceedings . 2017 edn, vol. 15, Australian Language Technology Association (ALTA), Melbourne Vic Australia, pp. 34-42, Australasian Language Technology Association Workshop 2017, Brisbane, Queensland, Australia, 6/12/17.

Leveraging linguistic resources for improving neural text classification. / Liu, Ming; Haffari, Gholamreza; Buntine, Wray et al.
Australasian Language Technology Association Workshop, ALTA 2017: 6–8 December 2017, Brisbane, Australia, Proceedings . ed. / Jojo Sze-Meng Wong; Gholamreza Haffari. Vol. 15 2017. ed. Melbourne Vic Australia: Australian Language Technology Association (ALTA), 2017. p. 34-42.

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

TY - GEN

T1 - Leveraging linguistic resources for improving neural text classification

AU - Liu, Ming

AU - Haffari, Gholamreza

AU - Buntine, Wray

AU - Ananda-Rajah, Michelle

N1 - Conference code: 15th

PY - 2017

Y1 - 2017

N2 - This paper presents a deep linguistic attentional framework which incorporates word level concept information into neural classification models. While learning neural classification models often requires a large amount of labelled data, linguistic concept information can be obtained from external knowledge, such as pre-trained word embeddings, WordNet for common text and MetaMap for biomedical text. We explore two different ways of incorporating word level concept annotations, and show that leveraging concept annotation scan boost the model performance and reduce the need for large amounts of labelled data. Experiments on various data sets validate the effectiveness of the proposed method.

AB - This paper presents a deep linguistic attentional framework which incorporates word level concept information into neural classification models. While learning neural classification models often requires a large amount of labelled data, linguistic concept information can be obtained from external knowledge, such as pre-trained word embeddings, WordNet for common text and MetaMap for biomedical text. We explore two different ways of incorporating word level concept annotations, and show that leveraging concept annotation scan boost the model performance and reduce the need for large amounts of labelled data. Experiments on various data sets validate the effectiveness of the proposed method.

M3 - Conference Paper

VL - 15

SP - 34

EP - 42

BT - Australasian Language Technology Association Workshop, ALTA 2017

A2 - Wong, Jojo Sze-Meng

A2 - Haffari, Gholamreza

PB - Australian Language Technology Association (ALTA)

CY - Melbourne Vic Australia

T2 - Australasian Language Technology Association Workshop 2017

Y2 - 6 December 2017 through 8 December 2017

ER -

Liu M, Haffari G, Buntine W, Ananda-Rajah M. Leveraging linguistic resources for improving neural text classification. In Wong JSM, Haffari G, editors, Australasian Language Technology Association Workshop, ALTA 2017: 6–8 December 2017, Brisbane, Australia, Proceedings . 2017 ed. Vol. 15. Melbourne Vic Australia: Australian Language Technology Association (ALTA). 2017. p. 34-42