Distributed latent variable models of lexical co-occurrences

John Blitzer; Amir Globerson; Fernando Pereira

Distributed latent variable models of lexical co-occurrences

John Blitzer, Amir Globerson, Fernando Pereira

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

Abstract

Low-dimensional representations for lexical co-occurrence data have become increasingly important in alleviating the sparse data problem inherent in natural language processing tasks. This work presents a distributed latent variable model for inducing these low-dimensional representations. The model takes inspiration from both connectionist language models [1, 16] and latent variable models [13, 9]. We give results which show that the new model significantly improves both bigram and trigram models.

Original language	English
Title of host publication	AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics
Pages	25-32
Number of pages	8
State	Published - 2005
Externally published	Yes
Event	10th International Workshop on Artificial Intelligence and Statistics, AISTATS 2005 - Hastings, Christ Church, Barbados Duration: 6 Jan 2005 → 8 Jan 2005

Publication series

Name	AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics

Conference

Conference	10th International Workshop on Artificial Intelligence and Statistics, AISTATS 2005
Country/Territory	Barbados
City	Hastings, Christ Church
Period	6/01/05 → 8/01/05

Cite this

@inproceedings{ef53e94b529d461e8ef7219f15b6b4f3,

title = "Distributed latent variable models of lexical co-occurrences",

abstract = "Low-dimensional representations for lexical co-occurrence data have become increasingly important in alleviating the sparse data problem inherent in natural language processing tasks. This work presents a distributed latent variable model for inducing these low-dimensional representations. The model takes inspiration from both connectionist language models [1, 16] and latent variable models [13, 9]. We give results which show that the new model significantly improves both bigram and trigram models.",

author = "John Blitzer and Amir Globerson and Fernando Pereira",

year = "2005",

language = "אנגלית",

isbn = "097273581X",

series = "AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics",

pages = "25--32",

booktitle = "AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics",

note = "null ; Conference date: 06-01-2005 Through 08-01-2005",

}

Blitzer, J, Globerson, A & Pereira, F 2005, Distributed latent variable models of lexical co-occurrences. in AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics. AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics, pp. 25-32, 10th International Workshop on Artificial Intelligence and Statistics, AISTATS 2005, Hastings, Christ Church, Barbados, 6/01/05.

Distributed latent variable models of lexical co-occurrences. / Blitzer, John; Globerson, Amir; Pereira, Fernando.

AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics. 2005. p. 25-32 (AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics).

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

TY - GEN

T1 - Distributed latent variable models of lexical co-occurrences

AU - Blitzer, John

AU - Globerson, Amir

AU - Pereira, Fernando

PY - 2005

Y1 - 2005

N2 - Low-dimensional representations for lexical co-occurrence data have become increasingly important in alleviating the sparse data problem inherent in natural language processing tasks. This work presents a distributed latent variable model for inducing these low-dimensional representations. The model takes inspiration from both connectionist language models [1, 16] and latent variable models [13, 9]. We give results which show that the new model significantly improves both bigram and trigram models.

AB - Low-dimensional representations for lexical co-occurrence data have become increasingly important in alleviating the sparse data problem inherent in natural language processing tasks. This work presents a distributed latent variable model for inducing these low-dimensional representations. The model takes inspiration from both connectionist language models [1, 16] and latent variable models [13, 9]. We give results which show that the new model significantly improves both bigram and trigram models.

UR - http://www.scopus.com/inward/record.url?scp=84862620701&partnerID=8YFLogxK

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:84862620701

SN - 097273581X

SN - 9780972735810

T3 - AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics

SP - 25

EP - 32

BT - AISTATS 2005 - Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics

Y2 - 6 January 2005 through 8 January 2005

ER -

Distributed latent variable models of lexical co-occurrences

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this