Co-attention-Based Pairwise Learning for Author Name Disambiguation

Shenghui Wang; Qiuke Li; Rob Koopman

doi:10.1007/978-981-99-8088-8_21

Co-attention-Based Pairwise Learning for Author Name Disambiguation

Shenghui Wang^*
, Qiuke Li
, Rob Koopman

^*Corresponding author for this work

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

3 Citations (Scopus)

114 Downloads (Pure)

Abstract

Digital libraries face a pressing issue of author name ambiguity. This paper proposes a novel pairwise learning model for author name disambiguation, utilizing self-attention and co-attention mechanisms. The model integrates textual, discrete, and co-author attributes, amongst others, to capture comprehensive information from bibliographic records. It incorporates an optional random projection-based dimension reduction technique for efficiency to handle large datasets. The attention weight visualizations provide explanations for the model’s predictions. Our experiments on a substantial bibliographic catalogue repository validate the model’s effectiveness using accuracy, F1, and ROC AUC scores.

Original language	English
Title of host publication	Leveraging Generative Intelligence in Digital Libraries
Subtitle of host publication	Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings
Editors	Dion H. Goh, Shu-Jiun Chen, Suppawong Tuarob
Publisher	Springer
Pages	240-249
Number of pages	10
ISBN (Electronic)	978-981-99-8088-8
ISBN (Print)	978-981-99-8087-1
DOIs	https://doi.org/10.1007/978-981-99-8088-8_21
Publication status	Published - 30 Nov 2023
Event	25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023 - Taipei, Taiwan Duration: 4 Dec 2023 → 7 Dec 2023 Conference number: 25

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher	Springer
Volume	14458
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023
Abbreviated title	ICADL 2023
Country/Territory	Taiwan
City	Taipei
Period	4/12/23 → 7/12/23

Keywords

Attention mechanisms
Author name disambiguation
Explainable machine learning
Feature Integration
2024 OA procedure

Access to Document

10.1007/978-981-99-8088-8_21Licence: Taverne

ArticleFinal published version, 1.11 MBLicence: Taverne

Cite this

Wang, S., Li, Q., & Koopman, R. (2023). Co-attention-Based Pairwise Learning for Author Name Disambiguation. In D. H. Goh, S.-J. Chen, & S. Tuarob (Eds.), Leveraging Generative Intelligence in Digital Libraries: Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings (pp. 240-249). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14458). Springer. https://doi.org/10.1007/978-981-99-8088-8_21

Wang, Shenghui ; Li, Qiuke ; Koopman, Rob. / Co-attention-Based Pairwise Learning for Author Name Disambiguation. Leveraging Generative Intelligence in Digital Libraries: Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings. editor / Dion H. Goh ; Shu-Jiun Chen ; Suppawong Tuarob. Springer, 2023. pp. 240-249 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{efb5206bf08d45e988fba9497dee4ff9,

title = "Co-attention-Based Pairwise Learning for Author Name Disambiguation",

abstract = "Digital libraries face a pressing issue of author name ambiguity. This paper proposes a novel pairwise learning model for author name disambiguation, utilizing self-attention and co-attention mechanisms. The model integrates textual, discrete, and co-author attributes, amongst others, to capture comprehensive information from bibliographic records. It incorporates an optional random projection-based dimension reduction technique for efficiency to handle large datasets. The attention weight visualizations provide explanations for the model{\textquoteright}s predictions. Our experiments on a substantial bibliographic catalogue repository validate the model{\textquoteright}s effectiveness using accuracy, F1, and ROC AUC scores.",

keywords = "Attention mechanisms, Author name disambiguation, Explainable machine learning, Feature Integration, 2024 OA procedure",

author = "Shenghui Wang and Qiuke Li and Rob Koopman",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.; 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, ICADL 2023 ; Conference date: 04-12-2023 Through 07-12-2023",

year = "2023",

month = nov,

day = "30",

doi = "10.1007/978-981-99-8088-8\_21",

language = "English",

isbn = "978-981-99-8087-1",

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

publisher = "Springer",

pages = "240--249",

editor = "Goh, \{Dion H.\} and Shu-Jiun Chen and Suppawong Tuarob",

booktitle = "Leveraging Generative Intelligence in Digital Libraries",

address = "Germany",

}

Wang, S, Li, Q & Koopman, R 2023, Co-attention-Based Pairwise Learning for Author Name Disambiguation. in DH Goh, S-J Chen & S Tuarob (eds), Leveraging Generative Intelligence in Digital Libraries: Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14458, Springer, pp. 240-249, 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Taipei, Taiwan, 4/12/23. https://doi.org/10.1007/978-981-99-8088-8_21

Co-attention-Based Pairwise Learning for Author Name Disambiguation. / Wang, Shenghui; Li, Qiuke; Koopman, Rob.
Leveraging Generative Intelligence in Digital Libraries: Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings. ed. / Dion H. Goh; Shu-Jiun Chen; Suppawong Tuarob. Springer, 2023. p. 240-249 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14458).

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

TY - GEN

T1 - Co-attention-Based Pairwise Learning for Author Name Disambiguation

AU - Wang, Shenghui

AU - Li, Qiuke

AU - Koopman, Rob

N1 - Conference code: 25

PY - 2023/11/30

Y1 - 2023/11/30

N2 - Digital libraries face a pressing issue of author name ambiguity. This paper proposes a novel pairwise learning model for author name disambiguation, utilizing self-attention and co-attention mechanisms. The model integrates textual, discrete, and co-author attributes, amongst others, to capture comprehensive information from bibliographic records. It incorporates an optional random projection-based dimension reduction technique for efficiency to handle large datasets. The attention weight visualizations provide explanations for the model’s predictions. Our experiments on a substantial bibliographic catalogue repository validate the model’s effectiveness using accuracy, F1, and ROC AUC scores.

AB - Digital libraries face a pressing issue of author name ambiguity. This paper proposes a novel pairwise learning model for author name disambiguation, utilizing self-attention and co-attention mechanisms. The model integrates textual, discrete, and co-author attributes, amongst others, to capture comprehensive information from bibliographic records. It incorporates an optional random projection-based dimension reduction technique for efficiency to handle large datasets. The attention weight visualizations provide explanations for the model’s predictions. Our experiments on a substantial bibliographic catalogue repository validate the model’s effectiveness using accuracy, F1, and ROC AUC scores.

KW - Attention mechanisms

KW - Author name disambiguation

KW - Explainable machine learning

KW - Feature Integration

KW - 2024 OA procedure

UR - https://www.scopus.com/pages/publications/85180151651

U2 - 10.1007/978-981-99-8088-8_21

DO - 10.1007/978-981-99-8088-8_21

M3 - Conference contribution

AN - SCOPUS:85180151651

SN - 978-981-99-8087-1

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

SP - 240

EP - 249

BT - Leveraging Generative Intelligence in Digital Libraries

A2 - Goh, Dion H.

A2 - Chen, Shu-Jiun

A2 - Tuarob, Suppawong

PB - Springer

T2 - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023

Y2 - 4 December 2023 through 7 December 2023

ER -

Wang S, Li Q, Koopman R. Co-attention-Based Pairwise Learning for Author Name Disambiguation. In Goh DH, Chen SJ, Tuarob S, editors, Leveraging Generative Intelligence in Digital Libraries: Towards Human-Machine Collaboration - 25th International Conference on Asia-Pacific Digital Libraries, ICADL 2023, Proceedings. Springer. 2023. p. 240-249. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-981-99-8088-8_21