The inevitable rise of machine learning in malware analysis puts forward the need for human-understandable explanations of the learned results. We point out how the ontological representation of malware data provides a suitable language for the construction of such explanations. We then focus on possible methods that enable producing such explanations and we reflect on our experience with them in the context of the EMBER dataset.
A note on methods for explainable malware analysis
Cardillo F. A.;Debole F.;Straccia U.;
2025
Abstract
The inevitable rise of machine learning in malware analysis puts forward the need for human-understandable explanations of the learned results. We point out how the ontological representation of malware data provides a suitable language for the construction of such explanations. We then focus on possible methods that enable producing such explanations and we reflect on our experience with them in the context of the EMBER dataset.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
shields-3.pdf
accesso aperto
Descrizione: A Note on Methods for Explainable Malware Analysis
Tipologia:
Versione Editoriale (PDF)
Licenza:
Creative commons
Dimensione
761.54 kB
Formato
Adobe PDF
|
761.54 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
