research interests - chunhualiao/public-docs GitHub Wiki
The top 5 keywords that best reflect my research interests and areas of expertise are:
1. OpenMP and Parallel Programming
Many publications focus on OpenMP extensions, tools for data race detection, and parallel programming models. This includes work on:
- OpenMP directive semantics
- OpenMP-based benchmarks (e.g., DataRaceBench)
- Extensions for runtime adaptation and accelerator support
- Interactive educational tools for OpenMP
2. Large Language Models (LLMs) for Code Analysis
A significant recent emphasis is on applying LLMs in:
- Translating and comparing Fortran and C/C++ code
- Compiler optimization analysis and improvement (e.g., CompilerGPT)
- Data race detection using LLMs
- Automating documentation and migration tasks
3. High-Performance Computing (HPC)
The researcher is deeply involved in:
- Optimization and correctness of HPC applications
- FAIR AI practices in HPC environments
- GPU memory optimization and unified memory frameworks
- Benchmarks and tools tailored for HPC software correctness
4. Compiler Technologies
Extensive work is evident in:
- Source-to-source transformation (e.g., ROSE compiler)
- Training platforms for compiler development (e.g., FreeCompilerCamp)
- Analysis and transformation of compiler optimization reports
- Static and dynamic tools for correctness and performance
5. Program Analysis and Software Tools
The publications include contributions to:
- Ontology-based program analysis and call graph synthesis
- Static and dynamic analysis tools for concurrency errors
- Automated migration and refactoring tools
- Performance modeling and visualization tools
These keywords encompass a research trajectory that integrates traditional compiler/HPC expertise with state-of-the-art machine learning and natural language processing tools for practical software engineering and system-level challenges.