EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Data and context matter: towards generalizing AI-based software vulnerability detection

Rijha Safdar, Danyail Mateen, Syed Taha Ali, Umer Ashfaq, Wajahat Hussain

International Journal of Information Security · 20.08.2026 · 25 (5)

DOI: 10.1007/s10207-026-01282-6 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Information Security, online erschienen am 20.08.2026. Verfasst von Rijha Safdar und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 82 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

AI-based solutions demonstrate remarkable results in identifying vulnerabilities in software, but research has consistently found that this performance does not generalize to unseen codebases. In this paper, we specifically investigate the impact of model architecture, parameter configuration, and quality of training data on the ability of these systems to generalize. For this purpose, we introduce VulGate, a high quality state of the art dataset that mitigates the shortcomings of prior datasets, by removing mislabeled and duplicate samples, updating new vulnerabilities, incorporating additional metadata, integrating hard samples, and including dedicated test sets. We undertake a series of experiments to demonstrate that improved dataset diversity and quality substantially enhances vulnerability detection. We also introduce and benchmark multiple encoder-only and decoder-only models. We find that encoder-based models outperform other models in terms of accuracy and generalization. Our model achieves 6.8% improvement in recall on the benchmark BigVul dataset and outperforms others on unseen projects, demonstrating enhanced generalizability. Our results highlight the role of data quality and model selection in the development of robust vulnerability detection systems. Our findings suggest a direction for future systems with high cross-project effectiveness.

Zitieren

APA

Safdar, R., Mateen, D., Ali, S. T., Ashfaq, U., & Hussain, W. (2026). Data and context matter: towards generalizing AI-based software vulnerability detection. International Journal of Information Security, 25(5). https://doi.org/10.1007/s10207-026-01282-6

Vancouver

Safdar R, Mateen D, Ali ST, Ashfaq U, Hussain W. Data and context matter: towards generalizing AI-based software vulnerability detection. International Journal of Information Security. 2026;25(5). doi:10.1007/s10207-026-01282-6

BibTeX
@article{Safdar2026,
  author = {Safdar, Rijha and Mateen, Danyail and Ali, Syed Taha and Ashfaq, Umer and Hussain, Wajahat},
  title = {Data and context matter: towards generalizing AI-based software vulnerability detection},
  journal = {International Journal of Information Security},
  year = {2026},
  volume = {25},
  number = {5},
  doi = {10.1007/s10207-026-01282-6}
}
Kostenlos starten →