VSR News - feature

At this year's SaxFDM Tagung, our team members Jan Haas and Christoph Göpfert held a workshop titled "Natürlichsprachliche Abfrage forschungsdatenorientierter institutioneller Wissensgraphen mittels großer Sprachmodelle" in context of the HYP*MOL project. In the workshop, they showcased how large language models can translate natural-language queries into SPARQL requests, enabling an intuitive exploration of complex knowledge graphs. With this approach, the entry barrier to querying knowledge graphs is significantly lowered for researchers across all disciplines.

Additionally, they presented in two posters showcasing the Knowledge Base of Come2Data as well as a strategy for holistically integrated information management (GIIM), which is currently being implemented at TU Chemnitz to integrate distributed research information from heterogeneous data sources.

We are very happy to welcome Bai Li to the VSR group as a researcher contributing to the Across project.

We look forward to the valuable contributions he will make and will fully support his research as he embarks on this new chapter.

Our team member Verena Traubinger presented our work on user preferences regarding different variations of chatbot interface patterns at the 9th International Symposium on Chatbots and Human-centred AI. She is also part of the program committee for this year's symposium.

CONVERSATIONS is a yearly symposium where researchers from the fields of chatbots and human-centred AI come together and discuss their insights regarding the creation, evaluation, and use of conversational interfaces.

Our team members Lucas Schröder and Christoph Göpfert represented the data competence center Come2Data at this year's Tag des wissenschaftlichen Nachwuchs at TU Chemnitz. Their participation helped raise awareness of data literacy in general and promoted Come2Data's training and support services in the areas of data integration, data management, data analysis and data publication.

Last week, our team member Sebastian Heil attended the ENFIELD Workshop on Human-centric AI and the ENFIELD General Assembly in Bucharest. In his talk, he presented joint research with Prof. Steven Furnell and Dr. Xavier Carpent from the University of Nottingham on the human perception of trust in AI. The presentation shared early findings from a large-scale longitudinal study of AI discourse in UK media and outlined the next steps: a vignette study exploring trust in AI applications across four ENFIELD domains—Healthcare, Energy, Space, and Manufacturing. The workshop, excellently organized by Prof. Adina Magda Florea, Dr. Alexandru Sorici, and the team at University POLITEHNICA of Bucharest, provided a valuable forum for exchanging ideas on Human-centric and Trustworthy AI.

This year's MuC Conference was a very special event for the VSR team and the other professorships involved from our university, as it was held right here at TU Chemnitz.

When we say "special," we mean both a wonderful and an incredibly demanding experience. While regular attendees enjoyed a 3.5-day conference, our dedicated volunteers worked for five full days, including setup and teardown. For some team members, preparations began months in advance.

As the General Chair, Prof. Martin Gaedke, and as the Local Chair, Dr. Sebastian Heil, led one of Europe's largest HCI conferences. Together with two other professorships from TU Chemnitz, the VSR team succeeded in providing around 500 attendees with a well-planned, highly engaging, and thoroughly successful MuC conference in Chemnitz. We also owe a huge thank you to the many student volunteers who actively helped, including a number of bachelor’s, master's and PhD students from our Web Engineering program.

The feedback we received, both in person and on LinkedIn, was overwhelmingly positive. Not only was the excellent organization praised, but the central theme of "Digital Diversity" and the inspiring workshops and sessions also received high acclaim. And, as one person noted, "Chemnitz is a beautiful place, by the way."

Our team member Maheshika Hansamalee Walpola participated in the 2nd Conference on Research Data Infrastructure (CoRDI 2025) at RWTH Aachen University in Aachen. Organized by the National Research Data Infrastructure (NFDI), CoRDI is Germany’s flagship forum for research data infrastructures. Under the theme “RDM in Action”, the event brought together many participants from NFDI consortia, research initiatives, and international partners to exchange ideas on research data management and FAIR data practices.

On Monday, our team member Sebastian Heil co-organized the International Workshop on Secure, Trustworthy, and Robust AI (STRAI) at the 20th International Conference on Availability, Reliability and Security (ARES2025) in Ghent. The workshop was part of the TrustworthyAI pillar in ENFIELD – AI Network of Excellence and attracted submissions from both Cybersecurity and Safety research. Sebastian had the opportunity to chair the workshop session together with Prof. Georgios Spathoulas and to exchange ideas with many enthusiastic PhD students at the beginning of their research journey, as well as with highly experienced senior researchers. The lively discussions underlined the importance of building AI systems that are not only powerful but also secure, robust, and trustworthy. The STRAI workshop highlighted the growing community around Trustworthy AI within ENFIELD, fostering collaboration across disciplines and institutions.

Last week, our team member Sebastian Heil helped organize the ENFIELD AI Summer School hosted by Prof. Geza Nemeth in Budapest at the Budapest University of Technology and Economics. He was Program Co-Chair together with Prof. Adina Magda Florea and invited speaker in the Trustworthy AI sessions. In his talk, he introduced the concept of Trustworthy AI, highlighting the fundamentals of trust and trustworthiness, with a particular focus on the seven key requirements outlined by the EU High-Level Expert Group on AI. He also presented methods and tools for building trustworthy AI systems, such as Explainable AI techniques and model cards, and offered insights into three current research directions pursued within the ENFIELD – AI Network of Excellence. Participants engaged hands-on session assessing the trustworthiness of their own AI systems. The ENFIELD Summer School brought together students and speakers from diverse backgrounds to explore AI from multiple disciplinary and industrial perspectives. VSR will organize the next ENFIELD AI Summer School in 2026, so for more information please visit the ENFIELD Website and social media channels.

We are very happy to welcome Sandra Schaftner to the VSR group as a researcher contributing to the Across project.

We look forward to the valuable contributions she will make and will fully support her research as she embarks on this new chapter.