Sebastian Simon
Leipzig University
Chair of Software Systems
Augustusplatz 10
04109 Leipzig
| Room: Phone: E-mail: | P635 (Paulinum) +49 (0) 341 97 323 92 ssimon@informatik.uni-leipzig.de |
Research Interest
- Configurable Software Systems
- Configuration Dependencies
- Retrieval-Augmented Generation
- Large Language Models
Courses
- WS 24/25 Softwaretechnikpraktikum (Bachelor)
- SS 2024 Software Engineering für KI-Systeme (Master)
- WS 23/24 Softwaretechnikpraktikum (Bachelor)
- SS 2023: Software Engineering für KI-Systeme (Master)
- WS 22/23: Softwaretechnikpraktikum (Bachelor)
- WS 21/22: Softwaretechnikpraktikum (Bachelor)
Supervised Theses
- Leveraging Configuration Network to Automatically Resolve Configuration Errors
- Automatic Identification of Configuration Errors from Stack Overflow
- Ermittlung von Konfigurationsoptionen im Source Code mit Focus auf Machine Learning Bibliotheken in Python
- Tracking Configuration Options in Source Code Focusing on Java Frameworks
- Literaturanalyse zu Entwicklungsansätzen und -herausforderungen intelligenter Softwaresysteme in der Praxis
- Klassifikation von Konfigurationsabhängigkeiten zwischen verschiedenen Technologien
Publications
2024
- Alina Mailach, Sebastian Simon, Johannes Dorn, and Norbert Siegmund. Practitioners‘ Discussions on Building LLM-based Applications for Production. 2024. arXiv e-prints, arXiv–2411.
- Sebastian Simon, Alina Mailach, Johannes Dorn, and Norbert Siegmund. A Methodology for Evaluating RAG Systems: A Case Study On Configuration Dependency Validation. 2024. arXiv preprint arXiv:2410.08801.
2023
- Sebastian Simon, Nicolai Ruckel, Norbert Siegmund. CfgNet: A Framework for Tracking Equality-Based Configuration Dependencies Across a Software Project. IEEE Transactions on Software Engineering (TSE), 2023. DOI: https://doi.org/10.1109/TSE.2023.3274349
- Sebastian Simon, Nikolay Kolyada, Christopher Akiki, Martin Potthast, Benno Stein, and Norbert Siegmund. Exploring Hyperparameter Usage and Tuning in Machine Learning Research. In 2nd International Conference on AI Engineering – Software Engineering for AI (CAIN), May, 2023. CAIN. DOI: 10.1109/CAIN58948.2023.00016.
Talks/Conferences
- ICSE 2024 (Journal-first Track): CfgNet: A Framework for Tracking Equality-based Configuration Dependencies Across a Software Project (April 14 – April 20, 2024, Lisbon, Portugal)
- FOSD 2024: Detecting and Extracting Configuration Dependencies from Stack Overflow Posts using Large Language Models (April 09 – April 12, 2024, Eindhoven, Netherlands)
- CAIN 2023: Exploring Hyperparameter Usage and Tuning in Machine Learning Research (upcoming, May 15-16, 2023, Online)
- FOSD 2023: Learning Configuration Dependencies Across the Technology Stack of Open-Source Software Projects (March 27-31, 2023, Ulm, Germany)
- FOSD 2022: Evolution of the Configuration Space in Modern Software Development (March 29 – April 01, 2022, Vienna, Austria)
- FOSD 2021: CfgNet Detecting Dependency Conflicts of Configurations (April 13-15, 2021, Online)