On Efficient Retrieval of Top Similarity Vectors

Shulong Tan; Zhixin Zhou; Zhaozhuo Xu; Ping Li

doi:10.18653/v1/d19-1527

On Efficient Retrieval of Top Similarity Vectors

Shulong Tan
, Zhixin Zhou
, Zhaozhuo Xu
, Ping Li

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

25 Citations (Scopus)

Abstract

Retrieval of relevant vectors produced by representation learning critically influences the efficiency in natural language processing (NLP) tasks. In this paper we demonstrate an efficient method for searching vectors via a typical nonmetric matching function: inner product. Our method, which constructs an approximate Inner Product Delaunay Graph (IPDG) for top-1Maximum Inner Product Search (MIPS), transforms retrieving the most suitable latent vectors into a graph search problem with great benefits of efficiency. Experiments on data representations learned for different machine learning tasks verify the outperforming effectiveness and efficiency of the proposed IPDG.

Original language	English
Title of host publication	Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Publisher	Association for Computational Linguistics
Pages	5236–5246
ISBN (Print)	9781950737901
DOIs	https://doi.org/10.18653/v1/d19-1527
Publication status	Published - Nov 2019
Externally published	Yes
Event	2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019) - Asia World Expo , Hong Kong, China Duration: 3 Nov 2019 → 7 Nov 2019 https://www.emnlp-ijcnlp2019.org/

Publication series

Name	EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference

Conference

Conference	2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019)
Place	Hong Kong, China
Period	3/11/19 → 7/11/19
Internet address	https://www.emnlp-ijcnlp2019.org/

Access Output

10.18653/v1/d19-1527

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Tan, S., Zhou, Z., Xu, Z., & Li, P. (2019). On Efficient Retrieval of Top Similarity Vectors. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) (pp. 5236–5246). (EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference). Association for Computational Linguistics. https://doi.org/10.18653/v1/d19-1527

Tan, Shulong ; Zhou, Zhixin ; Xu, Zhaozhuo et al. / On Efficient Retrieval of Top Similarity Vectors. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Association for Computational Linguistics, 2019. pp. 5236–5246 (EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference).

@inproceedings{30cd0adb792449e3bfb1e3aedca1e3cc,

title = "On Efficient Retrieval of Top Similarity Vectors",

abstract = "Retrieval of relevant vectors produced by representation learning critically influences the efficiency in natural language processing (NLP) tasks. In this paper we demonstrate an efficient method for searching vectors via a typical nonmetric matching function: inner product. Our method, which constructs an approximate Inner Product Delaunay Graph (IPDG) for top-1Maximum Inner Product Search (MIPS), transforms retrieving the most suitable latent vectors into a graph search problem with great benefits of efficiency. Experiments on data representations learned for different machine learning tasks verify the outperforming effectiveness and efficiency of the proposed IPDG. ",

author = "Shulong Tan and Zhixin Zhou and Zhaozhuo Xu and Ping Li",

year = "2019",

month = nov,

doi = "10.18653/v1/d19-1527",

language = "English",

isbn = "9781950737901",

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

publisher = "Association for Computational Linguistics",

pages = "5236–5246",

booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",

address = "United States",

note = "2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019) ; Conference date: 03-11-2019 Through 07-11-2019",

url = "https://www.emnlp-ijcnlp2019.org/",

}

Tan, S, Zhou, Z, Xu, Z & Li, P 2019, On Efficient Retrieval of Top Similarity Vectors. in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference, Association for Computational Linguistics, pp. 5236–5246, 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019), Hong Kong, China, 3/11/19. https://doi.org/10.18653/v1/d19-1527

On Efficient Retrieval of Top Similarity Vectors. / Tan, Shulong; Zhou, Zhixin; Xu, Zhaozhuo et al.
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Association for Computational Linguistics, 2019. p. 5236–5246 (EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on 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 - On Efficient Retrieval of Top Similarity Vectors

AU - Tan, Shulong

AU - Zhou, Zhixin

AU - Xu, Zhaozhuo

AU - Li, Ping

PY - 2019/11

Y1 - 2019/11

N2 - Retrieval of relevant vectors produced by representation learning critically influences the efficiency in natural language processing (NLP) tasks. In this paper we demonstrate an efficient method for searching vectors via a typical nonmetric matching function: inner product. Our method, which constructs an approximate Inner Product Delaunay Graph (IPDG) for top-1Maximum Inner Product Search (MIPS), transforms retrieving the most suitable latent vectors into a graph search problem with great benefits of efficiency. Experiments on data representations learned for different machine learning tasks verify the outperforming effectiveness and efficiency of the proposed IPDG.

AB - Retrieval of relevant vectors produced by representation learning critically influences the efficiency in natural language processing (NLP) tasks. In this paper we demonstrate an efficient method for searching vectors via a typical nonmetric matching function: inner product. Our method, which constructs an approximate Inner Product Delaunay Graph (IPDG) for top-1Maximum Inner Product Search (MIPS), transforms retrieving the most suitable latent vectors into a graph search problem with great benefits of efficiency. Experiments on data representations learned for different machine learning tasks verify the outperforming effectiveness and efficiency of the proposed IPDG.

UR - https://www.scopus.com/pages/publications/85075463883

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85075463883&origin=recordpage

U2 - 10.18653/v1/d19-1527

DO - 10.18653/v1/d19-1527

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

SN - 9781950737901

T3 - EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference

SP - 5236

EP - 5246

BT - Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

PB - Association for Computational Linguistics

T2 - 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019)

Y2 - 3 November 2019 through 7 November 2019

ER -

Tan S, Zhou Z, Xu Z, Li P. On Efficient Retrieval of Top Similarity Vectors. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Association for Computational Linguistics. 2019. p. 5236–5246. (EMNLP-IJCNLP - Conference on Empirical Methods in Natural Language Processing and International Joint Conference on Natural Language Processing, Proceedings of the Conference). doi: 10.18653/v1/d19-1527

On Efficient Retrieval of Top Similarity Vectors

Abstract

Publication series

Conference

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this