Research &
Publications.

Multi-document QA system for regulatory compliance using transformer models, applied to financial and healthcare specifications. One of the earliest published applications of transformer-based NLP to requirements compliance — predating widespread LLM adoption. Evaluated on real compliance documents.

• Automated compliance checking via transformer-based QA
• Multi-document reasoning across heterogeneous regulatory sources
• Domain-specific NLP evaluation on real financial and healthcare specs

Scalable NLP pipeline for automated requirements analysis — validated across 1,250+ industrial requirements from eight diverse domains. Open dataset and replication package released.

• Scalable pipeline for automated requirements quality analysis
• Industrial validation across 1,250+ requirements from 8 domains
• Open dataset and replication package for community use

Foundational ML approach for identifying requirements in natural language documents — evaluated on 22 real specifications across 8 domains. Now a frequently cited benchmark in the field.

• Machine learning approach to requirements demarcation
• Evaluated on 22 real-world specifications across 8 domains
• Established as a frequently cited benchmark in the field

One of the first published applications of RAG to requirements engineering in a live industrial setting — developed with automotive industry partners. Demonstrates how retrieval-augmented generation can bridge the gap between natural language requirements and executable test scenarios.

• First industrial application of RAG to requirements-driven test generation
• Co-developed and validated with automotive industry partners
• Bridges natural language requirements and executable test scenarios

Systematic approach to deriving executable test cases from natural language requirements using LLMs, evaluated in industrial contexts. Led by PhD student Fanyu Wang — published in ACM Transactions on Software Engineering and Methodology.

• Systematic LLM-based test case derivation from NL requirements
• Industrial evaluation across multiple real-world contexts
• Led by PhD student Fanyu Wang — collaborative supervision outcome

Pipeline for extracting structured requirements from unstructured user reviews at scale — applied to app store data across multiple domains.

• End-to-end pipeline from unstructured user feedback to structured requirements
• RAG-enhanced LLM approach for scalable elicitation
• Applied to app store reviews across multiple domains at scale