Table 16.
Interpretations of results.
| Component | Status | Interpretation |
|---|---|---|
| IOC Extraction | High accuracy (IPs, Domains) | Effective hybrid approach; improve entity disambiguation for malware terms |
| Extraction of IOC | High accuracy (IPs, Domains), | Improve entity disambiguation for malware terms using a hybrid approach |
| Threat Classification | Needs tuning | The classifier needs more datasets, and BERT performs best |
| Blockchain Integrity | Fully intact | A strong candidate to ensure report traceability and trust |
| Adaptive Learning | Requires more labeled data | Diverse samples, and a greater degree of effect to be effective |
| Visualization Metrics | Provides insight into trends | Demonstrates continuous learning’s scalability and value |
| Cross-Dataset Robustness Index (CRI) | BERT hitting a near-perfect 0.999 | Excellent cross-dataset stability |