Third NFDI Berlin-Brandenburg network meeting: AI as an enabler for science

Contribution List

7. Automated Workflows and Machine Learning for Materials Science Simulations

Dr Tilmann Hickel (BAM, MatWerk)

Tool Demos

Machine learning techniques in physics and materials science have revolutionized simulations and experimental analysis. Using these techniques to accurately predict, for example, material properties requires the manipulation and use of vast amounts of data. Manual processing and analysis quickly become impractical and error-prone, so the availability of automated workflows is critical to their...

4. Closing the loop in catalysis research

Dr Christoph Scheurer (FHI, NFDI4Cat)

Tool Demos

The mitigation of catalyst deactivation is of paramount importance for efficient technological processes. Due to the complexity of high-performance catalysts and their slow deterioration, it is challenging to study such processes and unravel their mechanisms. Self-driving Lab (SDL) approaches based on AI methods are required to conduct the necessary long-term experiments and drive them to...

5. From Queries to Conversations: Expanding Access to the MaRDI Knowledge Graph with a Chatbot Interface

Dr Tim Conrad (ZIB, MaRDI)

Tool Demos

The MaRDI Knowledge Graph contains structured information on mathematical research items such as mathematical papers, datasets, algorithms, and software. It can be explored in multiple ways: through a wiki-based website for intuitive browsing, as well as via an API and a SPARQL endpoint for more technical access. In this talk, we introduce a new chatbot interface that allows users to interact...

2. Multi-modal Large Language Models for Sustainability Challenges: Cases in Agriculture and Biodiversity

Prof. Masahiro Ryo (Leibniz Centre for Agricultural Landscape Research (ZALF))

Tool Demos

Masahiro Ryo explores the transformative potential of multimodal large language models (LLMs) for sustainability through two concrete examples: agriculture and biodiversity conservation. He illustrates efficiency gains in agriculture using AI techniques like YOLO, GPT-4, and T-Rex models, significantly simplifying agroecosystem predictions. In biodiversity conservation, he introduces modular...

9. NFDI4DS: Search and Explore Research Artifacts

Prof. Sonja Schimmler

Tool Demos

The talk gives an overview of the NFDI4DS Portal to search and explore research artifacts

6. NOMAD: a federated ecosystem for AI-ready data

Dr José A. Márquez Prieto (Humboldt Universität, FAIRmat)

Tool Demos

3. Scientific Competitions for AI-Driven Scholarly Document Processing

Prof. Georg Rehm (DFKI)

Keynote talk 2

This presentation will focus on a number of scientific competitions (also known as shared tasks) that we’ve been organising under the umbrella of the NFDI project NFDI4DataScience and AI (NFDI4DS). The talk will explain how scientific competitions work and what their overarching function in research is. The presentation will describe two iterations of the shared task Field of Research...

8. Wikidata: Wikimedia's knowledge graph in a world of generative AI

Alan Ang (Wikidata)

Keynote talk 1