A function-level defect localization method based on embedding technology

A positioning method, a function-level technology, applied in neural learning methods, electrical digital data processing, natural language data processing and other directions, can solve problems such as defect positioning effect, impact, insufficient semantic representation of defect reporting problems, etc. The effect of applying potential

Active Publication Date: 2022-03-18
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF13 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These technologies usually cannot represent the contextual semantics of the text well, resulting in insufficient semantic representation of the problem of defect reporting, thus affecting the effect of defect localization

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A function-level defect localization method based on embedding technology
  • A function-level defect localization method based on embedding technology
  • A function-level defect localization method based on embedding technology

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] Embodiments of the invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0027] Such as figure 1 As shown, the present invention proposes a function-level defect location method based on embedding technology, which aims to use word embedding and code embedding technology to effectively characterize the semantics of defect reports and function codes so as to effectively locate defects at the function granularity. Including the extraction of software defect reporting problem semantics and functional code functional semantics, the use of pre-trained models, model construction and prediction, and unbalanced class processing, etc.

[0028] figure 1 The workflow of the method proposed by the present invention is shown, including two stages: the s...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a function-level defect location method based on an embedding technology. The method uses an abstract syntax tree-based code embedding technology to represent function code function semantics, and uses word embedding technology to represent defect report problem semantics. Finally, a convolutional neural network is used to fuse the semantic features of functions and defect reports and predict suspicious functions associated with a given defect. In order to solve the problem of limited training data, the present invention proposes to use a pre-training model to perform feature representation on defect reports and codes; at the same time, it proposes to use random oversampling to deal with the problem of unbalanced number of class instances. Through experiments on three mainstream Java projects, it is found that when the recommended list is 10, the accuracy rate of the method proposed by the present invention can reach 12.5%-40%, showing great potential in the field of fine-grained defect location. It has great potential application value in software projects.

Description

technical field [0001] The invention relates to a function-level defect location method based on embedded technology, which belongs to the technical field of defect location and repair in the software development process. Background technique [0002] Software defect location is an important link in the process of software development and maintenance, and plays an irreplaceable role in ensuring and improving the quality of software products. In the past few decades, defect localization technology has been widely concerned and researched by the academic community, and is also considered by software development practitioners in the industry and open source communities to be the most valuable type of software technology. Defect location technology usually locates suspicious codes containing software defects based on program dynamic execution information or static text information, and the granularity of location may be source code files, functions, or code lines, etc. Among va...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06F11/36G06F40/289G06F40/30G06N3/04G06N3/08
CPCG06F11/3608G06F11/3684G06F11/3688G06F40/289G06F40/30G06N3/08G06N3/045
Inventor 邹卫琴李恩铭房春荣张静宣
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap