Vulnerability detection method based on random detection algorithm and information aggregation

Abstract

The invention discloses a vulnerability detection method based on a random detection algorithm and information aggregation. The method comprises the following steps: establishing an information set; decompiling; establishing a binary training set establishing a vulnerability detection model; judging abnormal points by the vulnerability detection model; bug handling. Network data flow information in each module of software is acquired, screened, organized and integrated for multiple times through the information aggregation module to obtain a high-purity binary training set, a vulnerability detection model is established by matching with the effects of a radial basis function neural network and a recurrent neural network, and through machine learning, the detection capability of the model on possible security vulnerabilities in binary software is continuously improved, and meanwhile, a random forest algorithm is adopted, so that the extraction accuracy of the function and the corresponding feature vector of the to-be-detected binary program is improved, and the purpose of rapid and accurate vulnerability detection is achieved.

Description

technical field [0001] The invention relates to the field of loophole detection, in particular to a loophole detection method based on a random detection algorithm and information aggregation. Background technique [0002] With the rapid development of computer information technology, cyberspace security is becoming more and more important, because the network incidents caused by software vulnerabilities are on the rise, which poses a great threat to cyberspace security. Vulnerability analysis technology has become a research hotspot in the field of network security. [0003] Vulnerability detection is an important branch of vulnerability analysis technology. Most of them start from existing vulnerabilities, find similar vulnerabilities that have a homologous relationship with them, and even discover undisclosed vulnerabilities based on the characteristics of vulnerability codes. However, when the existing vulnerability detection methods perform feature extraction, the numbe...

AI Technical Summary

Problems solved by technology

However, when the existing vulnerability detection methods perform feature extraction, the number of samples is large and complex, resulting in low efficiency of the vulnerability detection model, which affects the accuracy and pertinence of vulnerability detection.

Images