STAR Lab

Introduction

STAR Lab stands for Software Testing and Analysis Research Lab. Research done in STAR Lab brings safety and reliability of software to the software industry and helps users trust software they use. We develop automated techniques for software trustworthiness, reliablity, and security. We also validate and apply such techniques on real-world software systems.

Projects

We are working on the following projects:

  • Test generation for detecting bugs on conventional software
  • Fuzzing for detecting security vulnerabilites on AI software systems
  • Safety Testing of autonomous driving systems
  • Using NLP techniques to solve software testing tasks

We always encourage finding new problems and welcome new topics.

If you join STAR lab, you will be working on either one of these projects or any new topic that you want to explore.

For MS, MS-PhD, and PhD programs

As a graduate student, you will be a primary member of a research project (one of the mentioned projects above or something new). You will perform research by doing a literature survey on your topic, defining a research problem, developing a technique that addresses such a problem, and evaluating the technique. Finally, you will write a research paper of your project and publish it in a software engineering conference or journal, preferrably in a top-tier one such as ICSE, FSE, ISSTA, ASE, TSE, and TOSEM.

For UNIST undergraduate internship program

As an undergraduate student, you can choose either to work on assisting experiments of an existing research project or to study research basics in the mentioned projects above such as in safety testing of autonomous driving systems and fuzzing. The goal of doing an internship in our lab is to build your research experience in software testing and train your research abilities to eventually work on your own research project.

Preferred Skills

  • Strong programming skills at least in one language. Knowing Java or Python is a plus.
  • Experience on auto-driving simulation tools such as Carla ans Autoware
  • Experience on program analysis tools such as Soot, ASM, and LLVM

These are preferred skills, not required. We value your intest and willingness the most.

If you are interested, please send an email to Mijung at mijungk at unist dot ac dot kr.