dessertlab/cti-to-mitre-with-nlp

Implements multi-model classification of unstructured cybersecurity text into MITRE ATT&CK techniques using both traditional ML (SVM, Random Forest) and deep learning approaches (LSTM, CNN, fine-tuned SecBERT). The pipeline automatically generates training datasets from MITRE ATT&CK and CAPEC knowledge bases in STIX format, and includes evaluation scripts for analyzing model performance on real-world threat intelligence documents from APT groups like FIN6 and Wizard Spider.

No commits in the last 6 months.