PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain

Emmanuele Chersoni; Chu-ren Huang

doi:10.1145/3442442.3451387

PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain

Department of Language Science and Technology

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

4 Citations (Scopus)

Abstract

In this contribution, we describe the systems presented by the PolyU CBS Team at the second Shared Task on Learning Semantic Similarities for the Financial Domain (FinSim-2), where participating teams had to identify the right hypernyms for a list of target terms from the financial domain.

For this task, we ran our classification experiments with several distributional, string-based, and Transformer features. Our results show that a simple logistic regression classifier, when trained on a combination of word embeddings, semantic and string similarity metrics and BERT-derived probabilities, achieves a strong performance (above 90%) in financial hypernymy detection.

Original language	English
Title of host publication	WWW '21: Companion Proceedings of the Web Conference 2021
Editors	Jure Leskovec, Marko Grobelnik, Marc Najork, Jie Tang, Leila Zia
Publisher	Association for Computing Machinery (ACM)
Pages	316-319
ISBN (Electronic)	978-1-4503-8313-4
DOIs	https://doi.org/10.1145/3442442.3451387
Publication status	Published - 3 Jun 2021
Event	The First Workshop on Financial Technology on the Web - Online Duration: 14 Apr 2021 → 15 Apr 2021

Conference

Conference	The First Workshop on Financial Technology on the Web
Abbreviated title	FinWeb
Period	14/04/21 → 15/04/21

More information

10.1145/3442442.3451387

http://hdl.handle.net/10397/90384

Cite this

Chersoni, E., & Huang, C. (2021). PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain. In J. Leskovec, M. Grobelnik, M. Najork, J. Tang, & L. Zia (Eds.), WWW '21: Companion Proceedings of the Web Conference 2021 (pp. 316-319). Association for Computing Machinery (ACM). https://doi.org/10.1145/3442442.3451387

Chersoni, Emmanuele ; Huang, Chu-ren. / PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain. WWW '21: Companion Proceedings of the Web Conference 2021. editor / Jure Leskovec ; Marko Grobelnik ; Marc Najork ; Jie Tang ; Leila Zia. Association for Computing Machinery (ACM), 2021. pp. 316-319

@inproceedings{314956a50e604fbd9ec0025bbf8375c7,

title = "PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain",

abstract = "In this contribution, we describe the systems presented by the PolyU CBS Team at the second Shared Task on Learning Semantic Similarities for the Financial Domain (FinSim-2), where participating teams had to identify the right hypernyms for a list of target terms from the financial domain. For this task, we ran our classification experiments with several distributional, string-based, and Transformer features. Our results show that a simple logistic regression classifier, when trained on a combination of word embeddings, semantic and string similarity metrics and BERT-derived probabilities, achieves a strong performance (above 90\%) in financial hypernymy detection. ",

author = "Emmanuele Chersoni and Chu-ren Huang",

year = "2021",

month = jun,

day = "3",

doi = "10.1145/3442442.3451387",

language = "English",

pages = "316--319",

editor = "Jure Leskovec and Marko Grobelnik and Marc Najork and Jie Tang and Leila Zia",

booktitle = "WWW '21: Companion Proceedings of the Web Conference 2021",

publisher = "Association for Computing Machinery (ACM)",

address = "United States",

note = "The First Workshop on Financial Technology on the Web, FinWeb ; Conference date: 14-04-2021 Through 15-04-2021",

}

Chersoni, E & Huang, C 2021, PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain. in J Leskovec, M Grobelnik, M Najork, J Tang & L Zia (eds), WWW '21: Companion Proceedings of the Web Conference 2021. Association for Computing Machinery (ACM), pp. 316-319, The First Workshop on Financial Technology on the Web, 14/04/21. https://doi.org/10.1145/3442442.3451387

PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain. / Chersoni, Emmanuele ; Huang, Chu-ren.
WWW '21: Companion Proceedings of the Web Conference 2021. ed. / Jure Leskovec; Marko Grobelnik; Marc Najork; Jie Tang; Leila Zia. Association for Computing Machinery (ACM), 2021. p. 316-319.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain

AU - Chersoni, Emmanuele

AU - Huang, Chu-ren

PY - 2021/6/3

Y1 - 2021/6/3

N2 - In this contribution, we describe the systems presented by the PolyU CBS Team at the second Shared Task on Learning Semantic Similarities for the Financial Domain (FinSim-2), where participating teams had to identify the right hypernyms for a list of target terms from the financial domain. For this task, we ran our classification experiments with several distributional, string-based, and Transformer features. Our results show that a simple logistic regression classifier, when trained on a combination of word embeddings, semantic and string similarity metrics and BERT-derived probabilities, achieves a strong performance (above 90%) in financial hypernymy detection.

AB - In this contribution, we describe the systems presented by the PolyU CBS Team at the second Shared Task on Learning Semantic Similarities for the Financial Domain (FinSim-2), where participating teams had to identify the right hypernyms for a list of target terms from the financial domain. For this task, we ran our classification experiments with several distributional, string-based, and Transformer features. Our results show that a simple logistic regression classifier, when trained on a combination of word embeddings, semantic and string similarity metrics and BERT-derived probabilities, achieves a strong performance (above 90%) in financial hypernymy detection.

U2 - 10.1145/3442442.3451387

DO - 10.1145/3442442.3451387

M3 - Conference article published in proceeding or book

SP - 316

EP - 319

BT - WWW '21: Companion Proceedings of the Web Conference 2021

A2 - Leskovec, Jure

A2 - Grobelnik, Marko

A2 - Najork, Marc

A2 - Tang, Jie

A2 - Zia, Leila

PB - Association for Computing Machinery (ACM)

T2 - The First Workshop on Financial Technology on the Web

Y2 - 14 April 2021 through 15 April 2021

ER -

Chersoni E , Huang C. PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain. In Leskovec J, Grobelnik M, Najork M, Tang J, Zia L, editors, WWW '21: Companion Proceedings of the Web Conference 2021. Association for Computing Machinery (ACM). 2021. p. 316-319 doi: 10.1145/3442442.3451387

PolyU-CBS at the FinSim-2 Task: Combining Distributional, String-Based and Transformers-Based Features for Hypernymy Detection in the Financial Domain

Abstract

Conference

More information

Fingerprint

Cite this