Future of science: Discover digital transformation in Bielefeld!

Future of science: Discover digital transformation in Bielefeld!

The annual conference “Digital Humanities in German-speaking countries” took place in Bielefeld from March 3rd to 7th, 2025. This event, which is dedicated to the exchange about digital methods in the humanities, reached an important point in the discussion about new technologies in science with various lectures and performances.

The opening was by Professor Dr. Mark Dingemanse, a linguist at Radboud University Nijmegen, on March 4th at 6 p.m. His lecture entitled “Why are LLMs so irresistible?” was dedicated to the fascination that comes from AI systems. He looks at their historical use for interpretations and decisions. Dingemanse is also co-developer of the first global index for open source AI systems and leads the Futures of Language research group.

Openness and collaboration in research

On March 7th at 5 p.m. Dr. Mareike König, deputy director of the German Historical Institute in Paris, with her lecture “Done – for now: Unfinishedness as an epistemic value in the digital humanities”. She argues that the continued development of digital tools is a strength in humanities research. Both speakers are committed to the principle of open science. König runs the science blog portal de.hypotheses.org, while Dingemanse makes technologies testable.

The main lectures were accessible to members of Bielefeld University in the Audimax without prior registration and were also broadcast online. Their topics contributed to a broader discussion about the digital humanities and its challenges and opportunities.

Interdisciplinary approaches and knowledge graphs

A central topic of the conference was the use of knowledge graphs, network-like structures that map information and its semantic relationships. These graphs enable researchers to link information from different sources in a structured manner. They not only promote collaboration between interdisciplinary teams, but also help to identify connections and gaps between different research projects. By combining human and machine intelligence, knowledge graphs can support the generation of new hypotheses and significantly increase the efficiency of knowledge transfer.

The Open Research Knowledge Graph (ORKG) is an important resource that organizes scientific publications in a structured format. It is curated semi-automatically, with responsibility ultimately remaining with the researchers. The ORKG presents research problems, analysis methods, results and conclusions in a way that is understandable for both humans and machines. The ORKG Ask, a search and exploration system based on it, allows a targeted literature search and the synthetic provision of answers to research questions. There are currently 76 million scientific articles indexed in this semantic database.

The outlook on future technologies

Generative AI and large language models, combined with knowledge graphs, are key technologies for the future of scientific work. However, language models currently struggle to accurately answer specific scientific questions because they do not always have the latest research results. Nevertheless, integrating knowledge graphs into these technologies could significantly increase the accuracy and relevance of answering complex questions.

The conference also featured public events, such as a dance performance by Gaya Bommer Yemini, which implemented the conference's theme in an artistic way. These performances took place on March 3rd, 5th and 6th and tickets were still available. A workshop titled “Mind and Data” discussed the future of education in this digital era; Admission was free, but registration was required.

Overall, the annual conference shows how important interdisciplinary collaboration and the use of innovative technologies in research are in order to meet the challenges of digital transformation in the humanities.