S3M: Siamese Stack (Trace) Similarity Measure

@article{Khvorov2021S3MSS,
  title={S3M: Siamese Stack (Trace) Similarity Measure},
  author={Aleksandr Khvorov and Roman Vasiliev and George A. Chernishev and Irving Muller Rodrigues and Dmitrij V. Koznov and Nikita Povarov},
  journal={2021 IEEE/ACM 18th International Conference on Mining Software Repositories (MSR)},
  year={2021},
  pages={266-270}
}
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… 

Figures and Tables from this paper

Aggregation of Stack Trace Similarities for Crash Report Deduplication
TLDR
This paper proposes a new approach to solve the deduplication task and reports on its use on the real-world data from JetBrains, a leading developer of IDEs and other software.
FaST : A linear time stack trace alignment heuristic for crash report deduplication
TLDR
This paper proposes FaST, a novel sequence alignment method that computes the similarity score between two stack traces in linear time and evaluates FaST and five competing methods on four datasets from open-source projects using ranking and binary metrics.
DapStep: Deep Assignee Prediction for Stack Trace Error rePresentation
TLDR
This work reformulates this task as a ranking problem and proposes new deep learning models to solve it, based on a bidirectional recurrent neural network with attention and on a convolutional neural network, with the weights of the models optimized using a ranking loss function.

References

SHOWING 1-10 OF 28 REFERENCES
TraceSim: a method for calculating stack trace similarity
TLDR
TraceSim is described — a novel approach to this problem which combines TF-IDF, Levenshtein distance, and machine learning to construct a similarity metric which has been implemented inside an industrial-grade report triaging system.
A Soft Alignment Model for Bug Deduplication
TLDR
A novel deep learning network based on soft-attention alignment to improve duplicate bug report detection is presented, which outperforms state-of-the-art systems and strong baselines in different scenarios.
Finding Duplicates of Your Yet Unwritten Bug Report
  • Johannes Lerch, M. Mezini
  • Computer Science
    2013 17th European Conference on Software Maintenance and Reengineering
  • 2013
TLDR
This work proposes an approach that only uses stack traces and their structure as input to machine-learning algorithms for detecting bug-report duplicates, and shows that this approach performs as good as state-of-the-art techniques, but without requiring the whole text corpus of a bug report to be available.
DURFEX: A Feature Extraction Technique for Efficient Detection of Duplicate Bug Reports
TLDR
This paper proposes a feature extraction technique that reduces the feature size and yet retains the information that is most critical for the classification of duplicate bug reports, and outperforms the approach that uses distinct function names, while significantly reducing the processing time.
DeepTriage: Exploring the Effectiveness of Deep Learning for Bug Triaging
TLDR
This research proposes a novel bug report representation using a deep bidirectional recurrent neural network with attention (DBRNN-A) that learns the syntactic and semantic features from long word sequences in an unsupervised manner.
ReBucket: A method for clustering duplicate crash reports based on call stack similarity
TLDR
ReBucket is proposed, a method for clustering crash reports based on call stack matching that measures the similarities of call stacks in crash reports and then assigns the reports to appropriate buckets based on the similarity values.
Automatically Identifying Known Software Problems
TLDR
This paper exploits function call stacks as highly structured symptoms of a certain class of problems, including crashes, hangs, and traps, and proposes and evaluates algorithms for efficiently and accurately matching call stacks by a weighted metric of the similarity of their function names.
Train One Get One Free: Partially Supervised Neural Network for Bug Report Duplicate Detection and Clustering
TLDR
A neural architecture that can jointly detect if two bug reports are duplicates, and aggregate them into latent topics that can outperform state-of-the-art methods for duplicate classification on both cases, and learn meaningful latent clusters without additional supervision.
Revisiting the Performance Evaluation of Automated Approaches for the Retrieval of Duplicate Issue Reports
TLDR
It is shown that the classical evaluation tends to overestimate the performance of automated approaches for retrieving duplicate issue reports, and a realistic evaluation is proposed using all the reports that are available in the ITS of a software project to improve the performance.
The Unreasonable Effectiveness of Traditional Information Retrieval in Crash Report Deduplication
TLDR
Off-the-shelf information retrieval techniques, that were not designed to be used with crash reports, outperform other techniques which are specifically designed for the task of crash bucketing at realistic industrial scales in this paper.
...
...