Better Not to Use Vulnerability’s Reference for Exploitability Prediction

Abstract

About half of all exploit codes will become available within about two weeks of the release date of its vulnerability. However, 80% of the released vulnerabilities are never exploited. Since putting the same effort to eliminate all vulnerabilities can be somewhat wasteful, software companies usually use different methods to assess which vulnerability is more serious and needs an immediate patch. Recently, there have been some attempts to use machine learning techniques to predict a vulnerability’s exploitability. In doing so, a vulnerability’s related URL, called its reference, is commonly used as a machine learning algorithm’s feature. However, we found that some references contained proof-of-concept codes. In this paper, we analyzed all references in the National Vulnerability Database and found that 46,202 of them contained such codes. We compared prediction performances between feature matrix with and without reference information. Experimental results showed that test sets that used references containing proof-of-concept codes had better prediction performance than ones that used references without such codes. Even though the difference is not huge, it is clear that references having answer information contributed to the prediction performance, which is not desirable. Thus, it is better not to use reference information to predict vulnerability exploitation.

A Multiclass Approach to Estimating Software Vulnerability Severity Rating with Statistical and Word Embedding Methods

“…Few of the existing vulnerabilities are exploited by attackers [18]. This rate is about 20% of all software vulnerabilities [7]. This situation clearly shows us the advantage of being able to anticipate the deficits that merit focus.…”

“…CVE lists are used by many academic and empirical studies [5,6]. Half of the exploits that took place occurred within two weeks of posting the vulnerabilities [7]. This shows the importance of quickly analyzing security vulnerabilities and accurately determining their severity.…”

“…Although different systems have been proposed, the Common Vulnerability Scoring System (CVSS), which is officially used by NVD, has been accepted by all IT professionals and has become the standard in this field. The system, which is constantly updated, is currently available in 2.0 and 3.1 versions [7][8][9][10]. The main goal of CVSS is to create a common structure that accurately determines the seriousness and impact of security vulnerabilities [9].…”

Yazılım Güvenlik Açığı Veri Tabanları

“…Yang vd. (Yang et al 2020), istismar kodlarının yaklaşık yarısının güvenlik açığının ilanından iki hafta içerisinde kullanıldığı açıklamıştır. Bunun yanında ilan edilen açıların sadece %20'sinin istismara maruz kaldığı belirtilmektedir.…”

Efficient Algorithm for Providing Live Vulnerability Assessment in Corporate Network Environment

“…Furthermore, due to the ongoing Covid-19 pandemic [1] and related increased use of internet services, the cybersecurity issues have the potential to affect a much larger, than in previous years, part of the human population. Consequently, the identification and prioritization of vulnerabilities becomes a critical issue for a company that offers internet services [1,2].…”