基于自注意力机制和多特征提取的重复缺陷报告检测模型
DOI:
作者:
作者单位:

西南民族大学 计算机系统国家民委重点实验室

作者简介:

通讯作者:

中图分类号:

TP 311.5

基金项目:

国家自然科学基金资助项目(No.61502401, No.12050410248)、四川省科技计划项目(No.2021YFH0120)、西南民族大学中央高校基本科研业务费专项资金(No.2020YYXS59)。


Duplicate bug report detection model based on self-attention mechanism and multiple features extraction
Author:
Affiliation:

Southwest Minzu University,The Key Laboratory for Computer Systems of State Ethnic Affairs Commission

Fund Project:

Supported by the National Natural Science Foundation of China(No.61502401, No.12050410248), Sichuan Science and Technology Program(No.2021YFH0120 ), Fundamental Research Funds for the Central Universities, Southwest Minzu University (No.2020YYXS59).

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对重复缺陷报告检测研究中存在语义长距离依赖以及缺陷报告特征的单一性问题,提出一种基于自注意力机制和多特征提取的重复缺陷报告检测模型。引入自注意力机制捕获缺陷报告文本序列内部的语义关联性,从而动态计算上下文语义向量进行语义分析,解决长距离依赖问题;利用隐含狄利克雷分布算法捕获缺陷报告文本的主题特征,同时针对缺陷报告的类别信息,构建一种特征提取网络计算类别差异特征;最后基于三类特征向量进行综合检测。实验结果表明,该模型实现了更优的检测性能。

    Abstract:

    A duplicate bug report detection model based on self-attention mechanism and multiple features extraction was proposed to solve the problems of semantic long distance dependence and the singleness of bug report features in the current research on duplicate bug report detection. The self-attention mechanism was introduced to capture the semantic relevance within the bug report text sequence, calculate the contextual semantic vector in a dynamic way for semantic analysis, and solve the problem of long-distance dependence. The latent Dirichlet allocation algorithm was used to capture the topic characteristics of the bug report text, and a feature extraction network was constructed to calculate category difference features for category information of the bug report at the same time. Finally, comprehensive detection was performed based on three kinds of feature vectors. The experimental results show that the model achieves better detection performance.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2021-04-13
  • 最后修改日期:2021-05-31
  • 录用日期:2021-05-31
  • 在线发布日期:
  • 出版日期: