Sentence embedding and fine-tuning to automatically identify duplicate bugs

Haruna Isotani; Hironori Washizaki; Yoshiaki Fukazawa; Tsutomu Nomoto; Saori Ouji; Shinobu Saito

doi:10.3389/fcomp.2022.1032452

Sentence embedding and fine-tuning to automatically identify duplicate bugs

Haruna Isotani, Hironori Washizaki^*, Yoshiaki Fukazawa, Tsutomu Nomoto, Saori Ouji, Shinobu Saito

^*Corresponding author for this work

School of Fundamental Science and Engineering

Research output: Contribution to journal › Article › peer-review

Abstract

Industrial software maintenance is critical but burdensome. Activities such as detecting duplicate bug reports are often performed manually. Herein an automated duplicate bug report detection system improves maintenance efficiency using vectorization of the contents and deep learning–based sentence embedding to calculate the similarity of the whole report from vectors of individual elements. Specifically, sentence embedding is realized using Sentence-BERT fine tuning. Additionally, its performance is experimentally compared to baseline methods to validate the proposed system. The proposed system detects duplicate bug reports more effectively than existing methods.

Original language	English
Article number	1032452
Journal	Frontiers in Computer Science
Volume	4
DOIs	https://doi.org/10.3389/fcomp.2022.1032452
Publication status	Published - 2023 Jan 19

Keywords

BERT
bug reports
duplicate detection
information retrieval
natural language processing
sentence embedding

ASJC Scopus subject areas

Computer Science (miscellaneous)
Human-Computer Interaction
Computer Vision and Pattern Recognition
Computer Science Applications

Access to Document

10.3389/fcomp.2022.1032452

Cite this

@article{bf5a1bd4e51946ac965b0290c10e984b,

title = "Sentence embedding and fine-tuning to automatically identify duplicate bugs",

abstract = "Industrial software maintenance is critical but burdensome. Activities such as detecting duplicate bug reports are often performed manually. Herein an automated duplicate bug report detection system improves maintenance efficiency using vectorization of the contents and deep learning–based sentence embedding to calculate the similarity of the whole report from vectors of individual elements. Specifically, sentence embedding is realized using Sentence-BERT fine tuning. Additionally, its performance is experimentally compared to baseline methods to validate the proposed system. The proposed system detects duplicate bug reports more effectively than existing methods.",

keywords = "BERT, bug reports, duplicate detection, information retrieval, natural language processing, sentence embedding",

author = "Haruna Isotani and Hironori Washizaki and Yoshiaki Fukazawa and Tsutomu Nomoto and Saori Ouji and Shinobu Saito",

note = "Funding Information: We thank Yuta Koizumi, Zhenyu Xu, Akira Kanamaru, Tatsuka Muramoto, and Naohiko Tsuda for their support. We also would like to thank the reviewers for their insightful comments and suggestions. Publisher Copyright: Copyright {\textcopyright} 2023 Isotani, Washizaki, Fukazawa, Nomoto, Ouji and Saito.",

year = "2023",

month = jan,

day = "19",

doi = "10.3389/fcomp.2022.1032452",

language = "English",

volume = "4",

journal = "Frontiers in Computer Science",

issn = "2624-9898",

publisher = "Frontiers Media S. A.",

}

TY - JOUR

T1 - Sentence embedding and fine-tuning to automatically identify duplicate bugs

AU - Isotani, Haruna

AU - Washizaki, Hironori

AU - Fukazawa, Yoshiaki

AU - Nomoto, Tsutomu

AU - Ouji, Saori

AU - Saito, Shinobu

N1 - Funding Information: We thank Yuta Koizumi, Zhenyu Xu, Akira Kanamaru, Tatsuka Muramoto, and Naohiko Tsuda for their support. We also would like to thank the reviewers for their insightful comments and suggestions. Publisher Copyright: Copyright © 2023 Isotani, Washizaki, Fukazawa, Nomoto, Ouji and Saito.

PY - 2023/1/19

Y1 - 2023/1/19

N2 - Industrial software maintenance is critical but burdensome. Activities such as detecting duplicate bug reports are often performed manually. Herein an automated duplicate bug report detection system improves maintenance efficiency using vectorization of the contents and deep learning–based sentence embedding to calculate the similarity of the whole report from vectors of individual elements. Specifically, sentence embedding is realized using Sentence-BERT fine tuning. Additionally, its performance is experimentally compared to baseline methods to validate the proposed system. The proposed system detects duplicate bug reports more effectively than existing methods.

AB - Industrial software maintenance is critical but burdensome. Activities such as detecting duplicate bug reports are often performed manually. Herein an automated duplicate bug report detection system improves maintenance efficiency using vectorization of the contents and deep learning–based sentence embedding to calculate the similarity of the whole report from vectors of individual elements. Specifically, sentence embedding is realized using Sentence-BERT fine tuning. Additionally, its performance is experimentally compared to baseline methods to validate the proposed system. The proposed system detects duplicate bug reports more effectively than existing methods.

KW - BERT

KW - bug reports

KW - duplicate detection

KW - information retrieval

KW - natural language processing

KW - sentence embedding

UR - http://www.scopus.com/inward/record.url?scp=85147262669&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85147262669&partnerID=8YFLogxK

U2 - 10.3389/fcomp.2022.1032452

DO - 10.3389/fcomp.2022.1032452

M3 - Article

AN - SCOPUS:85147262669

SN - 2624-9898

VL - 4

JO - Frontiers in Computer Science

JF - Frontiers in Computer Science

M1 - 1032452

ER -

Sentence embedding and fine-tuning to automatically identify duplicate bugs

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this