Shellcode and annotation generation method based on dual learning

A technology of code generation and duality, which is applied in the computer field to achieve the effect of improving work efficiency, performance and accuracy

Pending Publication Date: 2022-05-27
NANTONG UNIVERSITY
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

According to the survey, there is only one study on shellcode code generation, which is the first attempt to use neural machine translation to solve this generation task. In addition, there is no study on generating annotations for shellcode.

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
  • Shellcode and annotation generation method based on dual learning
  • Shellcode and annotation generation method based on dual learning
  • Shellcode and annotation generation method based on dual learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0037] see figure 1 As shown, the present invention provides a method for generating shellcode codes and annotations based on dual learning, which specifically includes the following contents:

[0038] (1) Collect shellcode corpora from 2000 to 2020 on shellstorm and Exploit Database, and set the format to . Since the method of the present invention only uses one model to learn dual tasks, it directly exchanges input and output by exchanging input and output. to transform the corpus and distinguish the two tasks by adding prefixes: for shellcode generation tasks, prefix their input data with "ShellCodeGen:"; for shellcode comment generation tasks, prefix their input data with "ShellCodeSum:". Table 1 shows the statistics of the corpus.

[0039] Table 1

[0040]

[0041]

[0042] Annotation length statistics

[0043]

[0044] (2) Considering the characteristics of shellcode generation and annotation generation tasks, they are formalized as dual tasks. Use shallow T...

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 provides a method for generating a shellcode and annotation based on dual learning. The method comprises the following steps: (1) collecting a shellcode corpus from a shellstorm and an Exploit Database; (2) two tasks of shellcode annotation generation and shellcode generation are formalized into a dual learning problem, the two tasks are distinguished by adding a prefix for input, and synchronous learning is performed by using a shallow Transform, so that knowledge sharing is realized, and the performance and generalization ability of a training model are improved; (3) proposing a normalization method AdjostQKNorm for improving self-attention calculation, so that Transform is enabled to adapt to a low-resource task; and (4) providing a rule-based repair component. The method has the advantages that software development and maintenance are facilitated, the generated high-quality annotations and codes can help to improve the working efficiency of developers, and then the software quality is improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method for generating shellcode codes and annotations based on dual learning. Background technique [0002] Sellcode is a small piece of code written in assembly language, malicious shellcode can perform DDoS attacks, data theft, or run malware against the target system. However, shellcode with good intentions can be used to identify vulnerabilities and then fix the system by fixing those vulnerabilities. Therefore, researchers hope to use shellcode to find security vulnerabilities in software, and shellcode code generation technology has become an active research topic. [0003] However, automatically generating shellcodes is a challenging task since they are usually written in assembly language. In addition, summarizing the functional description of shellcodes (i.e., generating comments for shellcodes) is also a challenge, since identifying the attack intent of these shel...

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): G06F8/41G06F8/73G06F40/169G06F40/237G06K9/62G06N3/04
CPCG06F8/447G06F8/73G06F40/169G06F40/237G06N3/047G06F18/22Y02D10/00
Inventor陈翔周彦琳杨光于池刘珂顾亚锋夏鸿崚胡新宇
OwnerNANTONG UNIVERSITY