RAG-generated functional descriptions of Android app features paired with CySecBERT achieve 92.89% accuracy and 96.69% recall, beating an LLM fusion baseline but not a feature-based baseline.
Droidecho: an in-depth dissection of malicious behaviors in android applications
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2025 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Enhancing Android Malware Detection with Retrieval-Augmented Generation
RAG-generated functional descriptions of Android app features paired with CySecBERT achieve 92.89% accuracy and 96.69% recall, beating an LLM fusion baseline but not a feature-based baseline.