Conceptual sentence embeddings

Yashen Wang; Heyan Huang; Chong Feng; Qiang Zhou; Jiahui Gu

doi:10.1007/978-3-319-39937-9_30

Conceptual sentence embeddings

Yashen Wang^*, Heyan Huang, Chong Feng, Qiang Zhou, Jiahui Gu

^*Corresponding author for this work

School of Computer Science and Technology

Beijing Institute of Technology

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

2 Citations (Scopus)

Abstract

Most sentence embedding models typically represent each sentence only using word surface, which makes these models indiscriminative for ubiquitous homonymy and polysemy. In order to enhance discriminativeness, we employ concept conceptualization model to assign associated concepts for each sentence in the text corpus, and learn conceptual sentence embedding (CSE). Hence, the sentence representations are more expressive than some widely-used document representation models such as latent topic models, especially for short text. In the experiments, we evaluate the CSE models on two tasks, text classification and information retrieval. The experimental results show that the proposed models outperform typical sentence embedding models.

Original language	English
Title of host publication	Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings
Editors	Jianliang Xu, Nan Zhang, Dexi Liu, Bin Cui, Xiang Lian
Publisher	Springer Verlag
Pages	390-401
Number of pages	12
ISBN (Print)	9783319399362
DOIs	https://doi.org/10.1007/978-3-319-39937-9_30
Publication status	Published - 2016
Event	17th International Conference on Web-Age Information Management, WAIM 2016 - Nanchang, China Duration: 3 Jun 2016 → 5 Jun 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9658
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	17th International Conference on Web-Age Information Management, WAIM 2016
Country/Territory	China
City	Nanchang
Period	3/06/16 → 5/06/16

Keywords

Conceptualization
Sentence embedding
Text representation

Access to Document

10.1007/978-3-319-39937-9_30

Cite this

Wang, Y., Huang, H., Feng, C., Zhou, Q., & Gu, J. (2016). Conceptual sentence embeddings. In J. Xu, N. Zhang, D. Liu, B. Cui, & X. Lian (Eds.), Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings (pp. 390-401). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9658). Springer Verlag. https://doi.org/10.1007/978-3-319-39937-9_30

Wang, Yashen ; Huang, Heyan ; Feng, Chong et al. / Conceptual sentence embeddings. Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings. editor / Jianliang Xu ; Nan Zhang ; Dexi Liu ; Bin Cui ; Xiang Lian. Springer Verlag, 2016. pp. 390-401 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{fecd61e76c0a404c938e63c1671f429c,

title = "Conceptual sentence embeddings",

abstract = "Most sentence embedding models typically represent each sentence only using word surface, which makes these models indiscriminative for ubiquitous homonymy and polysemy. In order to enhance discriminativeness, we employ concept conceptualization model to assign associated concepts for each sentence in the text corpus, and learn conceptual sentence embedding (CSE). Hence, the sentence representations are more expressive than some widely-used document representation models such as latent topic models, especially for short text. In the experiments, we evaluate the CSE models on two tasks, text classification and information retrieval. The experimental results show that the proposed models outperform typical sentence embedding models.",

keywords = "Conceptualization, Sentence embedding, Text representation",

author = "Yashen Wang and Heyan Huang and Chong Feng and Qiang Zhou and Jiahui Gu",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2016.; 17th International Conference on Web-Age Information Management, WAIM 2016 ; Conference date: 03-06-2016 Through 05-06-2016",

year = "2016",

doi = "10.1007/978-3-319-39937-9_30",

language = "English",

isbn = "9783319399362",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "390--401",

editor = "Jianliang Xu and Nan Zhang and Dexi Liu and Bin Cui and Xiang Lian",

booktitle = "Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings",

address = "Germany",

}

Wang, Y, Huang, H, Feng, C, Zhou, Q & Gu, J 2016, Conceptual sentence embeddings. in J Xu, N Zhang, D Liu, B Cui & X Lian (eds), Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9658, Springer Verlag, pp. 390-401, 17th International Conference on Web-Age Information Management, WAIM 2016, Nanchang, China, 3/06/16. https://doi.org/10.1007/978-3-319-39937-9_30

Conceptual sentence embeddings. / Wang, Yashen; Huang, Heyan; Feng, Chong et al.
Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings. ed. / Jianliang Xu; Nan Zhang; Dexi Liu; Bin Cui; Xiang Lian. Springer Verlag, 2016. p. 390-401 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9658).

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

TY - GEN

T1 - Conceptual sentence embeddings

AU - Wang, Yashen

AU - Huang, Heyan

AU - Feng, Chong

AU - Zhou, Qiang

AU - Gu, Jiahui

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2016.

PY - 2016

Y1 - 2016

N2 - Most sentence embedding models typically represent each sentence only using word surface, which makes these models indiscriminative for ubiquitous homonymy and polysemy. In order to enhance discriminativeness, we employ concept conceptualization model to assign associated concepts for each sentence in the text corpus, and learn conceptual sentence embedding (CSE). Hence, the sentence representations are more expressive than some widely-used document representation models such as latent topic models, especially for short text. In the experiments, we evaluate the CSE models on two tasks, text classification and information retrieval. The experimental results show that the proposed models outperform typical sentence embedding models.

AB - Most sentence embedding models typically represent each sentence only using word surface, which makes these models indiscriminative for ubiquitous homonymy and polysemy. In order to enhance discriminativeness, we employ concept conceptualization model to assign associated concepts for each sentence in the text corpus, and learn conceptual sentence embedding (CSE). Hence, the sentence representations are more expressive than some widely-used document representation models such as latent topic models, especially for short text. In the experiments, we evaluate the CSE models on two tasks, text classification and information retrieval. The experimental results show that the proposed models outperform typical sentence embedding models.

KW - Conceptualization

KW - Sentence embedding

KW - Text representation

UR - http://www.scopus.com/inward/record.url?scp=84976621475&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-39937-9_30

DO - 10.1007/978-3-319-39937-9_30

M3 - Conference contribution

AN - SCOPUS:84976621475

SN - 9783319399362

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 390

EP - 401

BT - Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings

A2 - Xu, Jianliang

A2 - Zhang, Nan

A2 - Liu, Dexi

A2 - Cui, Bin

A2 - Lian, Xiang

PB - Springer Verlag

T2 - 17th International Conference on Web-Age Information Management, WAIM 2016

Y2 - 3 June 2016 through 5 June 2016

ER -

Wang Y, Huang H, Feng C, Zhou Q, Gu J. Conceptual sentence embeddings. In Xu J, Zhang N, Liu D, Cui B, Lian X, editors, Web-Age Information Management - 17th International Conference, WAIM 2016, Proceedings. Springer Verlag. 2016. p. 390-401. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-39937-9_30

Conceptual sentence embeddings

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this