TY - GEN

T1 - S3M

T2 - 18th IEEE/ACM International Conference on Mining Software Repositories, MSR 2021

AU - Khvorov, Aleksandr

AU - Vasiliev, Roman

AU - Chernishev, George

AU - Rodrigues, Irving Muller

AU - Koznov, Dmitrij

AU - Povarov, Nikita

N1 - A. Khvorov, R. Vasiliev, G. Chernishev, I. M. Rodrigues, D. Koznov and N. Povarov, "S3M: Siamese Stack (Trace) Similarity Measure," 2021 IEEE/ACM 18th International Conference on Mining Software Repositories (MSR), 2021, pp. 266-270, doi: 10.1109/MSR52588.2021.00038.

PY - 2021/5

Y1 - 2021/5

N2 - Automatic crash reporting systems have become a de-facto standard in software development. These systems monitor target software, and if a crash occurs they send details to a backend application. Later on, these reports are aggregated and used in the development process to 1) understand whether it is a new or an existing issue, 2) assign these bugs to appropriate developers, and 3) gain a general overview of the application's bug landscape. The efficiency of report aggregation and subsequent operations heavily depends on the quality of the report similarity metric. However, a distinctive feature of this kind of report is that no textual input from the user (i.e., bug description) is available: it contains only stack trace information.In this paper, we present S3M ("extreme") - the first approach to computing stack trace similarity based on deep learning. It is based on a siamese architecture that uses a biLSTM encoder and a fully-connected classifier to compute similarity. Our experiments demonstrate the superiority of our approach over the state-of-the-art on both open-sourced data and a private JetBrains dataset. Additionally, we review the impact of stack trace trimming on the quality of the results.

AB - Automatic crash reporting systems have become a de-facto standard in software development. These systems monitor target software, and if a crash occurs they send details to a backend application. Later on, these reports are aggregated and used in the development process to 1) understand whether it is a new or an existing issue, 2) assign these bugs to appropriate developers, and 3) gain a general overview of the application's bug landscape. The efficiency of report aggregation and subsequent operations heavily depends on the quality of the report similarity metric. However, a distinctive feature of this kind of report is that no textual input from the user (i.e., bug description) is available: it contains only stack trace information.In this paper, we present S3M ("extreme") - the first approach to computing stack trace similarity based on deep learning. It is based on a siamese architecture that uses a biLSTM encoder and a fully-connected classifier to compute similarity. Our experiments demonstrate the superiority of our approach over the state-of-the-art on both open-sourced data and a private JetBrains dataset. Additionally, we review the impact of stack trace trimming on the quality of the results.

KW - Automatic Crash Reporting

KW - Crash Report

KW - Deduplication

KW - Deep Learning

KW - Stack Trace

UR - http://www.scopus.com/inward/record.url?scp=85113586719&partnerID=8YFLogxK

UR - https://www.mendeley.com/catalogue/6d5241b4-01cb-393b-b023-06eeeb134895/

U2 - 10.1109/MSR52588.2021.00038

DO - 10.1109/MSR52588.2021.00038

M3 - Conference contribution

AN - SCOPUS:85113586719

SN - 978-1-6654-2985-6

T3 - IEEE International Working Conference on Mining Software Repositories

SP - 266

EP - 270

BT - 2021 IEEE/ACM 18th International Conference on Mining Software Repositories, MSR 2021

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 17 May 2021 through 19 May 2021

ER -