| Autor: |
Amine Daoud, Mohamed, Ahmed Mokhtar Mostefaoui, Sid, Ouared, Abdelkader, Madani Meghazi, Hadj, Mebarek, Bendaoud, Bouguessa, Abdelkader, Ahmed, Hasan |
| Zdroj: |
Journal of Computer Languages; 20240101, Issue: Preprints |
| Abstrakt: |
Creating an intrusion detection system (IDS) is a prominent area of research that continuously draws attention from both scholars and practitioners who tirelessly innovate new solutions. The complexity of IDS naturally escalates alongside technological advancements, whether they are manually implemented within security infrastructures or elaborated upon in academic literature. However, accessing and comparing these IDS solutions requires sifting through a multitude of hypotheses presented in research papers, which is a laborious and error-prone endeavor. Consequently, many researchers encounter difficulties in replicating results or reanalyzing published IDSs. This challenge primarily arises due to the absence of a standardized process for elucidating IDS methodologies. In response, this paper advocates for a framework aimed at enhancing the reproducibility of IDS outcomes, thereby enabling their seamless reuse across diverse cybersecurity contexts, benefiting both end-users and experts alike. The proposed framework introduces a descriptive language for the precise specification of IDS descriptions. Additionally, a model repository facilitates the sharing and reusability of IDS configurations. Lastly, through a case study, we showcase the effectiveness of our framework in addressing challenges associated with data acquisition and knowledge organization and sharing. Our results demonstrate satisfactory prediction accuracy for configuration reuse and precise identification of reusable components. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
Full Text from ScienceDirect