Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Fault Prediction Method Based on Software Network Feature Learning

A technology of fault prediction and network characteristics, applied in software testing/debugging, error detection/correction, instruments, etc., can solve problems such as ignoring macroscopic integrity correlation, and achieve the effect of reducing manpower and time costs

Active Publication Date: 2020-05-26
BEIHANG UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0020] The present invention is aimed at the existing defect prediction mainly by analyzing the software code or the development process, using the statistical feature data on the file or class related to the software defect as the measurement element, ignoring the macro integrity of the software program to a certain extent, To deal with issues such as the correlation between local defects and surrounding program elements, complex network technology is introduced into defect prediction. By establishing a software network model and using complex network measurement parameters to design a set of measurement elements that can reflect the local and overall characteristics of defects in the software, And propose a prediction model optimization method based on dynamic prediction threshold filtering algorithm

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 Fault Prediction Method Based on Software Network Feature Learning
  • A Fault Prediction Method Based on Software Network Feature Learning
  • A Fault Prediction Method Based on Software Network Feature Learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be further described in detail and in-depth below in conjunction with the accompanying drawings.

[0040] A kind of fault prediction method based on software network feature learning of the present invention, such as figure 1 , as a whole consists of five steps:

[0041] Step 1, determine the target software, obtain its complete software source code, and obtain the defect information of each version.

[0042] Step 2, using static code scanning technology and defect fusion technology to build a software defect network. Such as figure 2 Shown is a schematic diagram of a software defect network built for a certain version of the TintBrowser software.

[0043] Step 3, on the basis of step 2, design a multi-layer structure metric element for internal features, local features, and global features of software defect network nodes, and cons...

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 fault prediction method based on software network feature learning, and belongs to the field of software complicated networks. The method comprises the steps that source codes and defect information of each version of target software are obtained to construct a software defect network; multi-layered structure metric elements aiming at internal features, local features andglobal features of software defect network nodes are designed to construct a defect prediction data set; preprocessing is conducted on a defect prediction data set by using a maximized minimum methodand a PCA dimensionality reduction method; the best defect prediction model suitable for the software is established by using prediction model preference method of a dynamic prediction threshold filtering algorithm, and defect prediction effects and scores are given. According to the method, a complicated network technology is introduced to the defect prediction, a set of matric elements capableof reflecting the local and global features of defects in the software is designed by using complicated network matric parameters, the most suitable defect prediction model of the target software canbe selected to perform defect and fault prediction on the target software, and manpower and time cost is reduced to the most extent.

Description

technical field [0001] The invention is applied to the field of complex software networks, and is a fault prediction method based on software network feature learning. Background technique [0002] With the rapid development of network and information technology in recent decades, software, as a carrier for computers to realize various functions and assist people in various activities, plays an important role in all walks of life in the world today. An efficient and safe software system is highly dependent on software quality, and software defects that affect software quality are the root causes of system errors, failures, crashes, and even disasters. Therefore, software defect prediction has become a hot research field in recent years. At present, carrying out a large number of software tests is an important means to improve software reliability, and the annual development costs for supporting software tests are hundreds of millions. However, with the increasing scale and...

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): G06F11/36
CPCG06F11/368G06F11/3688G06F11/3692G06F11/3696
Inventor 艾骏杨益文苏文翥王飞郭皓然邹卓良
Owner BEIHANG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products