Here we present projects that define themselves as associated CDLs or as a network of associated CDLs.
CDLs in the clusters of excellence
The Cluster of Excellence AIM joints forces from physics, chemistry, and structural biology to observe, understand, and control processes at the scale of atoms, molecules, and solids in order to access new functionalities. In all these disciplines, data science plays an important role for handling the large amounts of data and a continuing development of digitization is imperative. In these research fields, exploiting AI methods is a new and fruitful direction, to which members of the CoE are making important contributions, e.g., in analysis of x-ray diffraction data and cryo electron microscopy data of macromolecules, in the optimization of quantum algorithms or in quantum state characterization. The machine-learning activities within AIM are coordinated within a task force:
CUI: Task Force on Machine Learning
The CoE UWA creates a multidisciplinary environment aiming at understanding written artefacts as objects having multiple dimensions including their digital representations, materiality, textual content and historical context. The involvement of data science and AI in UWA extends from the use of established methods to analyse research data by researchers in natural sciences and the Humanities, to the development of novel methods in machine learning and pattern recognition by computer scientists. Annual relevant scientific events, both local and international, are organized to facilitate collaboration between different disciplines and motivate the development of novel data science and AI approaches for researching written artefacts with the aim of better understanding them. Furthermore, UWA aims to develop novel chemical and physical methods and implement them for investigating organic/biological and inorganic systems. These technologies and experiments have the potential to assist in understanding the materials used in the production, usage, and alteration of written artefacts, such as manuscripts, throughout time.
- Data-Linking: https://www.csmc.uni-hamburg.de/written-artefacts/research-fields/field-f.html
- Pattern Analysis Software Tools: https://www.csmc.uni-hamburg.de/publications/software.html
- Similarity Measurement of Visual Patterns in Written Artefacts: https://www.csmc.uni-hamburg.de/written-artefacts/research-fields/field-a/rfa05.html
Quantum Universe (QU)
The cluster of excellence Quantum Universe (QU) conducts research across 45 orders of magnitude, from exploring the structure of matter in particle physics to observing the universe in astrophysics, complemented by theoretical and mathematical research. As a result, there is a wide variety of advanced digital methods used in QU research, ranging from machine learning for data analysis and generation, to microsecond latency decision systems, and the integration of mathematical structures in deep learning. In many cases QU research pioneers digital methods for research applications, such in the development of anomaly detection techniques in the search for physics beyond the Standard Model.
Beyond international scientific collaborations like experiments at CERN, Quantum Universe research is embedded in national (PUNCH4NFDI, Consortia in ErUM-Data) and local (HCDS, CDCS, the Graduate School DASHH, ...) initiatives to connect AI research in fundamental physics. The cluster organises a data science basic lecture series and a networking seminar "Physics ❤ AI" to facilitate exchanges between researchers.
For more information, visit the cluster homepage (https://www.qu.uni-hamburg.de/) or contact the cooordinator of the clusters' data science platform, Prof. Kasieczka (gregor.kasieczka"AT"uni-hamburg.de).
Center for Data and Computing in Natural Sciences (CDCS)
As a pilot study for the HCDS, the CDCS is a joint facility of the University of Hamburg, the German Electron Synchrotron DESY and the Technical University of Hamburg. Located in the newly emerging Science City Bahrenfeld, it combines natural science research with state-of-the-art information technology. The CDCS is a project of the Landesforschungsförderung Hamburg X (funding code LFF-HHX-03) under the direction of Matthias Rarey. It comprises 4 CDLs and a Computational Core Unit, which serves as a blueprint for the Methodology Competence Center.
Project title: Computational Astro- and Particle Physics
Lead: Peter Schleper, Frank Gaede
Funding: CDCS, LFF Hamburg X
Project title: Computational Photon Science
Lead: Anton Barty
Funding: CDCS, LFF Hamburg X
Project title: Computational Systems Biology
Lead: Jan Baumbach, Kay Grünewald
Funding: CDCS, LFF Hamburg X
Project title: Computational Controls of Accelerators
Lead: Görschwin Fey, Holger Schlarb
Funding: CDCS, LFF Hamburg X
Center for Biomedical AI - bAIome
The Center for Biomedical AI - bAIome - at the University Medical Center Hamburg-Eppendorf (UKE) was founded in 2019 with the goal of integrating cutting-edge AI research into clinical routine, where it can contribute to a better understanding of diseases and their therapy. The affiliation of bAIome with UKE creates a unique opportunity for collaborations in sandboxes, safe spaces for continuous testing and improvement of novel AI approaches using clinical data. In a sandbox, deep medical data meets AI expertise and medical specialists. Together, they create deployable applications and develop new ideas for using the vast amount of available biomedical data. The sandbox environment is designed to protect patient data.
Working groups and institutes of the bAIome:
- IMSB (Medical System Biologie)
- ICNS (Computational Neuroscience)
- IPMI (Image Processing and Medical Imaging)
- IAM (Applied Medical Informatics)
- Systems Immunology
- Department of Excellence for Neural Information Processing
- Research IT
This collaborative project between computer scientists, climate scientists, and model developers focuses on two challenges: on the one hand, the increasing computational demands of climate research, and on the other hand, the bottlenecks that limit the efficient scalable use of advances in high-performance computing. The goals of this project are conceptual improvement and the development and application-oriented evaluation of innovative methods in the following areas: Climate Model Software Engineering, Big Data Visualization, and Data Management for Simulations and Observations.
Project title: High-Perfomance Computing and Data-Intensive Science
Lead: Andrea Lammert, Stephan Olbrich
Funding: DFG, Exzellenzcluster CLICCS
Duration: since 2019
In the SLM departments, the area of Digital Humanities (DH) is established as a methodological cross-sectional paradigm in research and teaching. It focuses on the technical and conceptual digitization of literary and linguistic objects and research practices.
Digital Discourse Analyses in the Sociology of Knowledge
The D-WISE project develops new computational analysis methods for the use of context-oriented embedding representations and a prototypical working environment as digital support for the Sociology of Knowledge Approach to Discourse (SKAD). The project investigates for which purposes, when, and how digital humanities procedures can be usefully integrated into qualitative discourse-analytical knowledge production.
Project title: D-WISE: Digitale Wissenssoziologische Diskursanalyse
Lead: Gertraud Koch, Chris Biemann
Duration: 2021 - 2024
Research Unit Data Linking
One of the central ideas of the Cluster of Excellence "Understanding Written Artifacts" is to combine methods from the Humanities with the methods of the Natural Sciences for the study of written artifacts. The materials of written artifacts are analyzed with new technologies and different types of data (e.g. X-ray data, spectral data) are generated. Automatic linking of these analysis data with humanities research findings and data from publications, AI-assisted annotation, and AI-based support for humanities research overall is the goal of the Research Unit Data Linking
ITMC Service & AI Lab (SAIL)
ITMC-SAIL is a cross-disciplinary lab in collaboration of the IT Management and Consulting Trustees and the Department of Informatics. The SAIL lab will foster research and co-innovation on AI-based natural language assistants in service processes, exploring new modes of human-AI collaboration in these work settings. Furthermore, the lab will research the management and monitoring of AI-based systems, a key challenge in IT management research. Core to the lab is the evaluation of AI-based approaches in real-life settings in organizations.
Computational Human Dynamics (CHD) – Multimodal Social Signal Processing of Dyadic and Group Interactions
The Computational Human Dynamics (CHD) Lab is based on an interdisciplinary collaboration between the Department of Informatics (esp. Image Processing, Language Technology, Signal Processing, Human-Computer Interaction, Socio-Technical Systems Design and Ethics in Information Technology) and the Department of Psychology (esp. Social Psychology, Industrial and Organizational Psychology, Educational Psychology) as well as the Communication and Media Sciences and the Educational Sciences at Universität Hamburg. In the CHD Lab, AI-based methods for multimodal processing of social signals and communication channels (esp. visual, verbal, and paralinguistic) are developed, analyzed, and evaluated for the investigation of dyadic interactions and group processes as well as the associated change dynamics. In addition to the technical development of AI methods, the CHD Lab also combines competencies in empirical data collection for training AI models and reflects on the epistemological, ethical and social implications of applying AI methods to better understand human interaction patterns.
Lead: Frank Steinicke, Nale Lehmann-Willenbrock
- LFF project on Mechanisms of Change in Dynamic Social Interactions
- SFB/Transregio "Crossmodal Learning"
- BMBF project Hybrid Intelligent Virtual Avatar/Assistent-Models to support (tele-)medical Consulting and Treatment
Initialized Research Partnerships
- Mechanisms of Change
Duration: since 2021