Automated Cyber Threat Identification (NLP)
Built a Django-based NLP system for cyber threat classification using TF-IDF and ML models. Implemented text preprocessing and feature extraction with Logistic Regression, SVM, and Naive Bayes models. Achieved 79.3% accuracy and ~85% F1-score by optimizing model performance through comparison and tuning. Developed REST APIs and model caching, reducing inference time to ~100–200ms.