University of Twente Student Theses

Login
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.

Tracking the Evolution : Uncovering Concept Drift in Vulnerabilities Descriptions Over Time

Wijngaarden, Hugo van (2025) Tracking the Evolution : Uncovering Concept Drift in Vulnerabilities Descriptions Over Time.

[img] PDF
1MB
Abstract:The field of cyber security is one that is constantly changing and evolving. A part of this field consists of Common Vulnerabilities and Exposures (CVEs) and their descriptions. The nature of this field makes it difficult to keep machine learning models that map CVEs to Common Weakness Enumerations (CWEs) up to date. To keep these models relevant, this paper addresses the problem of concept drift in vulnerability descriptions. By finding an optimal training window, this paper improves the training strategy of language models on the CVE dataset to better reflect the current cyber security landscape and allows for a more accurate mapping of CVEs to CWEs. Various time windows have been evaluated, in Which the models trained on the two years immediately preceding the test set gave the best results. With this approach, a system for maintaining model relevance over time is proposed. This methodology will allow for a more accurate dataset of CVEs mapped to CWEs to be used in detecting cyber security threats.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:https://purl.utwente.nl/essays/107285
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page