Logo des Repositoriums

Auflistung nach Schlagwort "Synthetic data"

1 - 2 von 2
  • Konferenzbeitrag
    Generation of Plausible Synthetic Data for Stego-Malware Detection for Inter-zone IACS Protocols
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Edeh, Natasha; Altschaffel, Robert; Waedt, Karl
    ISO/IEC 27002:2022 distinguishes between the following types of security controls corrective: preventive, detective and corrective. The focus of this paper is on the support for testing of detective security controls for Industrial Automation and Control Systems. More specifically we will only address the generation of synthetic data that can be used for the detection of selected, advanced detective security controls. The proposed approach will be justified, while a comprehensive validation of the effectiveness of the synthetic data is beyond the scope of this paper. This work aims to contribute to the comprehension and improvement of security measures in Industrial Automation and Control Systems by focusing on the development of synthetic data and its consequences for the identification of specific detective security controls.
  • Konferenzbeitrag
    Scenario-based Data Set Generation for Use in Digital Forensics: A Case Study
    (INFORMATIK 2024, 2024) Göbel, Thomas; Baier, Harald; Wolf, Dennis
    Digital forensics is a rapidly growing and highly relevant field of cybersecurity. In case of an incident, the subsequent digital forensic investigation and analysis shall reveal the respective digital evidence. However, although electronic devices and their data play a central role in each crime investigation, data sets to train experts or to validate tools are sparse. While manual data set generation is a time-consuming, elaborate and error-prone task, tool-based data synthesis is an excellent candidate for simplifying data generation and solving the data set gap problem. Synthetic data sets can be used, for example, to test and refine forensic tools and methods under controlled conditions. In addition, entirely new approaches can be explored. Several promising data synthesis frameworks for digital forensic data set creation have been published lately, the most recent of which is ForTrace, a freely available, community-driven data synthesis framework written in Python for generating digital forensic data sets. This paper shows how to apply ForTrace in a large-scale manner without human interaction. Our main goal is to show the usability of ForTrace and demonstrate its practicality and benefits for the digital forensic domain. We therefore provide a sample usage of ForTrace in two scenarios, namely a VeraCrypt and a malware use case, and present the definition of the corresponding configurations.