Optimization of Software Vulnerabilities patterns with the Meta-Heuristic Algorithms
Küçük Resim Yok
Tarih
2022
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
In order to ensure the development of secure software, it is essential to predict software vulnerabilities. Nevertheless, there can be considerable losses in case of an attack on an information system. Detecting a dangerous code, which can lead to severe unknown consequences, requires great effort. There is a strong need to devise meta-heuristic-based approaches to provide effective security and prevent vulnerabilities or mitigate them. The primary focus of studies on software vulnerability prediction models is to specify the best set of predictors that are related to the presence of vulnerabilities. However, the existing vulnerability detection methods suffer from coarse detection granularity and a bias toward global features or local features. The framework proposed in the present work improves optimization algorithms for the best set of optimized vulnerability patterns correlated for software vulnerabilities based on a clock- work memory mechanism. Using the proposed framework, we found vulnerable optimized patterns based on clock-work memory mechanism feature representation learning that directly. The effectiveness of the developed algorithm was further improved with the clock-work memory mechanism based on 6 open-source projects, such as LibTIFF, Pidgin, FFmpeg, LibPNG, Asteriks, and VLC media player datasets.
Açıklama
Anahtar Kelimeler
Bilgisayar Bilimleri, Yazılım Mühendisliği, Optimization, Clock-work mechanism, Meta- heuristic algorithm, Software vulnerability
Kaynak
Türk Doğa ve Fen Dergisi
WoS Q Değeri
Scopus Q Değeri
Cilt
11
Sayı
4
