Construction of a Security Vulnerability Identification System Based on Machine Learning

• Main Authors: Kebin Shi, Yonghui Dai, Jing Xu
• Format: Article
• Language: English
• Published: Hindawi Limited 2020-01-01
• Series: Journal of Sensors
• Online Access: http://dx.doi.org/10.1155/2020/7358692
• ISSN: 1687-725X; 1687-7268

In recent years, the frequent outbreak of information security incidents caused by information security vulnerabilities has brought huge losses to countries and enterprises. Therefore, the research related to information security vulnerability has attracted many scholars, especially the research on the identification of information security vulnerabilities. Although some organizations have established information description databases for information security vulnerabilities, the differences in their descriptions and understandings of vulnerabilities have increased the difficulty of information security precautions. This paper studies the construction of a security vulnerability identification system, summarizes the system requirements, and establishes a vulnerability text classifier based on machine learning. It introduces the word segmentation, feature extraction, classification, and verification processing of vulnerability description text. The contribution of this paper is mainly in two aspects: One is to standardize the unified description of vulnerability information, which lays a solid foundation for vulnerability analysis. The other is to explore the research methods of a vulnerability identification system for information security and establish a vulnerability text classifier based on machine learning, which can provide reference for the research of similar systems in the future.