Publikationen
SCOPE: A Proof-of-Concept Framework for Measuring Alignment in Data-Scarce Cybergrooming Domains
| Autor | Choi, Jeong-Eun; Fan, Shiying; Steinebach, Martin |
|---|---|
| Datum | 2026 |
| Art | Conference Paper |
| Abstrakt | Cybergrooming is a growing threat to minors, but research is constrained by limited real-world data. Synthetic datasets are a promising alternative, yet their validity is difficult to assess. This paper introduces SCOPE, a framework for evaluating alignment between synthetic and real-world datasets in privacy-sensitive domains. We motivate the need for such evaluation and demonstrate SCOPE's feasibility with a proofof-concept example covering feature selection and alignment measurement. Supplementary materials are available at: SCOPE GitHub repository. |
| Konferenz | Annual Computers, Software, and Applications Conference 2026 |
| Url | https://publica.fraunhofer.de/handle/publica/523325 |
Schwachstellen und Sicherheitslücken
Aus- und Weiterbildung
Informieren Sie sich über die Aus- und Weiterbildungsangebote der in ATHENE mitwirkenden Institutionen