2026
A Structured Workflow for Transforming Cyber Threat Intelligence into Computable Detection Patterns
A protocol for converting threat-report indicators into validated detection patterns using LLM extraction and graph-assisted component labeling.
Selected publications.
Journal articles, conference papers, and preprints on threat intelligence, security automation, LLM evaluation, and applied machine learning.
Google Scholar2026
Why Machines Misread Pedagogical Quality: Human–Machine Alignment in LLM-Based Pretest Question Evaluation
How rubric design and evaluation mode affect alignment between human and LLM judgments.
Read preprint2026
From IOCs to Regex: Automating CTI Operationalization for SOC with LLMs
Group-aware generation and multi-stage validation for turning threat indicators into detection patterns.
2025
Exploring Prompt Patterns for Effective Vulnerability Repair in Real-World Code by Large Language Models
Investigating contextual prompts and repair strategies for vulnerabilities in real-world code.
Read at NIST2024
Using LLMs to Automate Threat Intelligence Analysis Workflows in Security Operation Centers
An LLM agent for repetitive threat-report analysis tasks in security operations.
2023
Vehicle theft detection by generative adversarial networks on driving behavior
Learning driving-behavior representations with a convolutional LSTM GAN for vehicle-theft detection.