In my research, I investigate how information from multiple sources can be captured in a model-driven manner to reconstruct software architecture models. In particular, I address the format, language, and semantic heterogeneity challenges resulting from the independent use and loose coupling of components and different technologies and platforms. I develop knowledge representation models for other sources and integrate individual views of them into a unified architectural model.

In teaching, I am interested in software engineering and the development of quality assurance tools. Fully automated tools provide fast feedback to students, while manual tools support the tutor during correction. Besides the classical approaches of statistical program analysis, I am also interested in using machine learning in software engineering. To this end, I have developed several techniques based on retrieval-augmented generation to support software engineering, especially software engineering education.

As one of the first Open Source Educational Software Lab members, my task is to support students in developing open-source software for educational purposes. We have already completed various projects in the Educational Software Lab, including an Android app for general education, web apps for computer science education, and Python apps for physics education.

Research Focus

View-based reverse engineering and traceability of web and microservices systems:

• Objective: Transform and understand complex systems to make model-based quality predictions.
• Methodology: Dissect and reconstruct system architectures to gain insight into system behavior and interactions.
• Benefits: Improve system safety, scalability, and maintainability through better prediction of quality attributes.
• Impact: Provide theoretical advances and practical tools for optimizing the performance and reliability of modern software systems.

Teaching Focus

Development and integration of quality assurance methodologies in software engineering education.

• Automated: Provide immediate feedback to students, improving learning efficiency.
• Manual: Assist tutors during the assessment process, improving grading accuracy.
• Innovation: Use Retrieval-Augmented LLMs to assist tutors in software engineering courses.

Contact