News Archive

In industrial production, getting the sequence right is crucial. What sounds trivial at first poses a considerable challenge for sectors with complex planning processes. The automotive industry is just one of many examples. Parts from suppliers must arrive on time and a wide variety of model configurations need to be mass-produced. At the same time, innovation cycles are becoming ever shorter. An AI-supported planning platform that optimally controls the production process has yet to be introduced.

Saarland University’s new discipline, Computational Social Science, is investigating how social science questions can be answered using methods from computer science and novel data sources. An English-language lecture series in the coming winter term will offer insights into this field of research at the interface between the social sciences and computer science. The lectures will take place every Friday from 12.15 pm to 1.45 pm (Building E1 7, Room 3.23). The series begins on 16 October.

How can large AI language models be adapted to become more reliable, secure and customisable? How can we understand what is happening inside these language models? Marius Mosbach, new Professor of Language Science and Technology at Saarland University, seeks to address these questions in his research, shedding light on the crucial step that transforms a pre-trained language model into a specialized, reliable system.

AI can free up more time for lawyers to focus on their actual legal work: At the EDV-Gerichtstag conference, taking place September 23 – 25 on the Saarbrücken campus, the Institute for Legal Informatics at Saarland University will present AI tools for research, teaching, and student studies. Professor Georg Borges estimates that AI could take over as much as 70 to 80 percent of the effort involved in otherwise time-consuming tasks such as searching for legal precedents.

Large language models such as ChatGPT have become an integral part of everyday life for many people and businesses. They are being used, e.g., to gather information from multiple sources and to support complex planning processes. Yet these AI systems are not infallible and they continue to make mistakes – LLMs hallucinate and struggle to combine information in a logically rigorous manner. Computational linguist Michael Hahn has set out to strengthen the theoretical foundations of AI reasoning.

On 2 September at Saarland University’s Innovation Center (Building A2 1), the SOUNDS data project will demonstrate how satellite imagery and Google Maps reviews can provide insights into life across Saarland. These kinds of analyses could also support decision-making processes in Saarland. SOUNDS is receiving financial support for this work from the Saarland state government’s Transformation Fund.

The ddpp project (Decentralized Data Privacy Protocol) aims to make it easier for users to manage their privacy settings and to strengthen their digital sovereignty on the web. The Fraunhofer Institute for Experimental Software Engineering IESE, which is leading the consortium, and researchers at Saarland University are developing an open web protocol that will allow users to manage their privacy preferences across multiple services at a location of their choice.

On 5 August, Prof. Dr. Jörg Siekmann will celebrate his 85th birthday. A pioneer of the field of artificial intelligence in Germany, Siekmann served as a professor of computer science at Saarland University and director of the Deduction and Multi-Agent Systems Research Division at the German Research Centre for Artificial Intelligence (DFKI).