Software vulnerability detection using SFDMN deep learning model
by Gayatri Bajantri; C. Noorullah Shariff
International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC), Vol. 47, No. 4, 2024

Abstract: In recent years, various software systems have been most commonly utilised in different life and production fields. The software vulnerabilities are the major security issues that occur in software. Hence, this paper presents ShuffleFDeepMaxout network (SFDMN) for the detection of vulnerabilities in software. Initially, the source-level intermediate representation (SIR) function from the input software is extracted for the creation of semantic subgraph. Similarly, the code exposure features are extracted, where the extracted code exposure features and semantic subgraph is fed into the SFDMN for the detection of software vulnerabilities. The different deep learning models, such as deep maxout network (DMN) and ShuffleNet are integrated to design the SFDMN model for accurately detecting the vulnerabilities. Thus, the experimental outcomes proved that the SFDMN achieved superior performance by varying the training data with maximum specificity of 92.62%, sensitivity of 89.97%, and accuracy of 89.99% by using the MSR_20_Code_vulnerability_CSV_Dataset and Security-patches-dataset.

Online publication date: Mon, 18-Nov-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Ad Hoc and Ubiquitous Computing (IJAHUC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com