Improving topic coherence with latent feature word representations in MAP estimation for topic modeling

Dat Quoc Nguyen; Kairit Sirts; Mark Johnson

Improving topic coherence with latent feature word representations in MAP estimation for topic modeling

Dat Quoc Nguyen, Kairit Sirts, Mark Johnson

School of Computing

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

5 Citations (Scopus)

Abstract

Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature word vectors have been used to obtain high performance in many natural language processing (NLP) tasks. In this paper, we present a new approach by incorporating word vectors to directly optimize the maximum a posteriori (MAP) estimation in a topic model. Preliminary results show that the word vectors induced from the experimental corpus can be used to improve the assignments of topics to words.

Original language	English
Title of host publication	Proceedings of the Australasian Language Technology Association Workshop 2015
Editors	Ben Hachey, Kellie Webster
Place of Publication	Stroudsburg, PA
Publisher	Association for Computational Linguistics (ACL)
Pages	116-121
Number of pages	6
Volume	13
Publication status	Published - 1 Dec 2015
Event	Australasian Language Technology Association Workshop (13th : 2015) - Parramatta, Australia Duration: 8 Dec 2015 → 9 Dec 2015

Conference

Conference	Australasian Language Technology Association Workshop (13th : 2015)
Abbreviated title	ALTA 2015
Country/Territory	Australia
City	Parramatta
Period	8/12/15 → 9/12/15

Keywords

MAP estimation
LDA
topic model
word vectors
topic coherence

Access to Document

http://www.aclweb.org/anthology/U15-1014.pdf

Cite this

Nguyen, D. Q., Sirts, K., & Johnson, M. (2015). Improving topic coherence with latent feature word representations in MAP estimation for topic modeling. In B. Hachey, & K. Webster (Eds.), Proceedings of the Australasian Language Technology Association Workshop 2015 (Vol. 13, pp. 116-121). Association for Computational Linguistics (ACL). http://www.aclweb.org/anthology/U15-1014.pdf

@inproceedings{7c4e5e3728c34f43b298d96b49d7cbe1,

title = "Improving topic coherence with latent feature word representations in MAP estimation for topic modeling",

abstract = "Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature word vectors have been used to obtain high performance in many natural language processing (NLP) tasks. In this paper, we present a new approach by incorporating word vectors to directly optimize the maximum a posteriori (MAP) estimation in a topic model. Preliminary results show that the word vectors induced from the experimental corpus can be used to improve the assignments of topics to words.",

keywords = "MAP estimation, LDA, topic model, word vectors, topic coherence",

author = "Nguyen, {Dat Quoc} and Kairit Sirts and Mark Johnson",

year = "2015",

month = dec,

day = "1",

language = "English",

volume = "13",

pages = "116--121",

editor = "Ben Hachey and Kellie Webster",

booktitle = "Proceedings of the Australasian Language Technology Association Workshop 2015",

publisher = "Association for Computational Linguistics (ACL)",

note = "Australasian Language Technology Association Workshop (13th : 2015), ALTA 2015 ; Conference date: 08-12-2015 Through 09-12-2015",

}

Nguyen, DQ, Sirts, K & Johnson, M 2015, Improving topic coherence with latent feature word representations in MAP estimation for topic modeling. in B Hachey & K Webster (eds), Proceedings of the Australasian Language Technology Association Workshop 2015. vol. 13, Association for Computational Linguistics (ACL), Stroudsburg, PA, pp. 116-121, Australasian Language Technology Association Workshop (13th : 2015), Parramatta, New South Wales, Australia, 8/12/15. <http://www.aclweb.org/anthology/U15-1014.pdf>

Improving topic coherence with latent feature word representations in MAP estimation for topic modeling. / Nguyen, Dat Quoc; Sirts, Kairit; Johnson, Mark.
Proceedings of the Australasian Language Technology Association Workshop 2015. ed. / Ben Hachey; Kellie Webster. Vol. 13 Stroudsburg, PA: Association for Computational Linguistics (ACL), 2015. p. 116-121.

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

TY - GEN

T1 - Improving topic coherence with latent feature word representations in MAP estimation for topic modeling

AU - Nguyen, Dat Quoc

AU - Sirts, Kairit

AU - Johnson, Mark

PY - 2015/12/1

Y1 - 2015/12/1

N2 - Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature word vectors have been used to obtain high performance in many natural language processing (NLP) tasks. In this paper, we present a new approach by incorporating word vectors to directly optimize the maximum a posteriori (MAP) estimation in a topic model. Preliminary results show that the word vectors induced from the experimental corpus can be used to improve the assignments of topics to words.

AB - Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature word vectors have been used to obtain high performance in many natural language processing (NLP) tasks. In this paper, we present a new approach by incorporating word vectors to directly optimize the maximum a posteriori (MAP) estimation in a topic model. Preliminary results show that the word vectors induced from the experimental corpus can be used to improve the assignments of topics to words.

KW - MAP estimation

KW - LDA

KW - topic model

KW - word vectors

KW - topic coherence

M3 - Conference proceeding contribution

VL - 13

SP - 116

EP - 121

BT - Proceedings of the Australasian Language Technology Association Workshop 2015

A2 - Hachey, Ben

A2 - Webster, Kellie

PB - Association for Computational Linguistics (ACL)

CY - Stroudsburg, PA

T2 - Australasian Language Technology Association Workshop (13th : 2015)

Y2 - 8 December 2015 through 9 December 2015

ER -

Improving topic coherence with latent feature word representations in MAP estimation for topic modeling

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this