检索结果-内蒙古大学图书馆

SSRN 2022年

作者： Mondal, Saikat Roy, Banani Software Research Lab Department of Computer Science University of Saskatchewan Saskatoon Canada Interactive Software Engineering and Analytics Lab Department of Computer Science University of Saskatchewan Saskatoon Canada

software developers often submit questions to Stack Overflow to resolve their code-level problems. They include example code snippets with questions to explain the programming issues. However, such code snippets could not always reproduce the issues that prevent questions from receiving appropriate solutions. A previous study investigated the potential challenges that might prevent reproducibility and produced a catalog. However, it is unknown how the practitioners perceive the challenge catalog. Practitioners’ perspective is inevitable to validate those challenges and estimate their severity. This study first surveys 53 practitioners to – (a) see practitioners’ viewpoints on the agreement to those challenges, (b) find the potential impact of those challenges, and (c) determine and prioritize tool support needs. Survey results show that – (a) about 90% of the participants agree with the challenges, (b) missing an important part of code, too short code snippets, and the absence of required Class/Interface/Method severely prevent reproducibility, and (c) participants strongly recommend introducing tool support to promote reproducibility. Second, we extract nine code-based features (e.g., LOC, compilability) and build five machine learning models to predict issue reproducibility. Our models can predict issue reproducibility with 84.5% precision, 83.0% recall, 82.8% F1-score & 82.8% overall accuracy, which are highly promising. © 2022, The Authors. All rights reserved.

关键词： Surveys

来源：评论

学校读者我要写书评

暂无评论

Feature Transformation for Improved software Bug Detection Models

SSRN

引用

SSRN 2022年

作者： Cynthia, Shamse Tasnim Roy, Banani Mondal, Debajyoti Interactive Software Engineering and Analytics Lab Department of Computer Science University of Saskatchewan Saskatoon Canada Lab Department of Computer Science University of Saskatchewan Saskatoon Canada

Testing software is considered to be one of the most crucial phases in software development life cycle. A rich body of recent research explored ways to predict bugs in software artifacts using machine learning techniques. For a reliable and trustworthy prediction, it is crucial to consider the explainability aspects of such machine learning models. In this paper, we show how the feature transformation techniques may significantly improve the prediction accuracy and build confidence into the working of bug prediction models. We propose a novel approach for improved bug prediction that first extracts the features, then finds a weighted transformation of these features using a genetic algorithm that best separates bugs from non-bugs when plotted in a low-dimensional space, and finally, trains the machine learning model using the transformed dataset. In our experiment with real-life bug datasets, the random forest and k-nearest neighbor classifiers that leveraged feature transformation showed 4.25% improvement in recall values on an average of over 8 software systems when compared to the models built on original data. We also examine the generalizability of the feature transformation concept for multiclass classification tasks such as commit classification in software systems and find modest improvement in F1-scores (sometimes upto 3%). © 2022, The Authors. All rights reserved.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：