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

Vulnerability detection method and system based on self-supervised learning and multi-channel hypergraph neural network

A neural network and vulnerability detection technology, applied in the direction of neural learning methods, biological neural network models, neural architectures, etc., can solve the problems of poor detection of loopholes and insufficient use of high-order relationships in codes, so as to achieve clean data stream information, The effect of low false alarm rate and reduced time overhead

Active Publication Date: 2022-07-08
EAST CHINA NORMAL UNIV
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that the existing loophole detection method cannot fully use the high-order relationship of the code and the loophole detection effect is poor, the purpose of the present invention is to provide a method based on self-supervised learning and multi-channel hypergraph neural network

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
  • Vulnerability detection method and system based on self-supervised learning and multi-channel hypergraph neural network
  • Vulnerability detection method and system based on self-supervised learning and multi-channel hypergraph neural network
  • Vulnerability detection method and system based on self-supervised learning and multi-channel hypergraph neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0103] The specific process of this embodiment is as follows:

[0104] First, select the code dataset QEMU and FFmpeg:

[0105] For the selected code data, the following describes how to convert the code text, such as figure 2 shown:

[0106] (1) Preprocess the code to remove links, special characters, etc. in the code.

[0107] (2) Standardize the code variable names.

[0108] (3) Using the compiler tool Joern to generate abstract syntax tree (AST graph) and control flow graph (CFG graph) of function code.

[0109] (4) Then traverse the token sequence of the leaf nodes in the AST, obtain the data transfer relationship between the code tokens, and generate a data flow graph (DFG graph).

[0110] (5) Then traverse the token sequences in the leaf nodes in the AST, match with the code text, and generate a sequence relationship diagram (SRG diagram) representing the sequence relationship between the code tokens before and after.

[0111] (6) Then, according to the Linear Di...

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 vulnerability detection method based on self-supervised learning and multi-channel hypergraph neural network. Motifs representing high-order information are constructed for different channels, and the code sequence graph is sampled according to the motif to obtain a multi-channel code sequence hypergraph. The preprocessed code text data is trained as word vector representation using word2vec. Finally, the code sequence hypergraph and labels are used as training data to train the hypergraph neural network, and the node representation and the hypergraph representation are obtained by learning, and then the hypergraph representation is spliced, and the graph is classified by a single-layer perceptron. The invention simultaneously introduces self-supervised learning to make up for the loss of information between multiple channels, and integrates the mutual information between multiple channels through self-supervised learning, which has better interpretability and vulnerability detection effect. The present invention also provides a system for implementing the above method.

Description

technical field [0001] The invention belongs to the technical field of computer information security, and relates to a vulnerability detection method and system based on self-supervised learning and a multi-channel hypergraph neural network, in particular to a method of constructing function-level codes into hypergraphs, utilizing self-supervised learning and hypergraphs A neural network's method of determining whether code is vulnerable. Background technique [0002] In recent years, with the rapid development of computer software technology, a large amount of software has been developed, and there are hidden loopholes in the software. Incorrect programming habits of developers and insufficient software testing of testers have resulted in a large number of hidden bugs in the code that have not been discovered. Hackers can take advantage of hidden loopholes, destroy systems, steal data, and cause greater harm to enterprises and countries. Therefore, vulnerability detection...

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
IPC IPC(8): G06F21/57G06N3/04G06N3/08
CPCG06F21/577G06N3/04G06N3/08
Inventor 王骏王志远张伟
Owner EAST CHINA NORMAL 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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com