Digital Total - Computing & Data Science an der Universität Hamburg und in der Wissenschaftsmetropole Hamburg

Liste der Beiträge

70. 123 - DisKo – Das Diversitäts Korpus (123)

Poster

DisKo steht für Diversitäts-Korpus und ist ein literaturwissenschaftliches Projekt mit Digital-Humanities-Komponente. Mit Methoden des maschinellen Lernens trainieren wir einen Machine Learning Classifiers, der nicht nur weibliche, männliche und neutrale Genderrollen in literarischen Texten aus unterschiedlichen Epochen erkennt, sondern auch weniger binäre Genderzuschreibungen. Für den...

158. A Computational Approach to Magnesium Corrosion Engineering (109)

Poster

Small organic molecules constitute useful materials to modify the service environment of light metals, such as magnesium (Mg). As the lightest engineering metal, Mg is promising for advanced technologies that will tackle climate change through improved battery technologies and advanced transport applications. However, due to its high chemical reactivity, target applications require tailoring...

64. A Personalized Web Application to Predict Changes in Knee Pain among Patients with Osteoarthritis Participating at the GLA:D® program (92)

Poster

GLA:Dai (https://apps.cosy.bio/gladai) is a web application for predicting individualized changes in knee pain for patients considering enrollment with the GLA:D® program (Good Life with osteoArthritis in Denmark). The statistical model is based on self-reported patient information to predict the individual change in pain intensity (VAS scale 0 to 100) from before to after GLA:D® enrollment....

35. Advanced mathematical solutions for circulating cardiovascular biomolecules (156)

Poster

Cardiovascular biomolecules contain quantifiable information about pathophysiologic processes and are thus frequently used in modern medicine. However, the magnitude of targets, their biological collinearity and their time-dependent variation constitute challenges for conventional statistical models, so that new mathematical and computational solutions are needed.
Within our...

76. ahoi.digital - Die Allianz der Hamburger Hochschulen für Informatik (64)

Poster

ahoi.digital is a program conceived based on the specific recommendations of the Wissenschaftsrat (WR) following their 2016 evaluation of computer science departments in Hamburg. The WR advised that Hamburg's higher education institutions should establish a closer, institutionalized collaboration among their computer science divisions through a cooperative platform. ahoi.digital was hence...

130. AI for Happiness: Pedestrian Path Generation through Agent-Based Simulations with Deep Reinforcement Learning (131)

Poster

AI integration in Smart Cities, primarily through agent-based simulations, holds transformative potential for understanding and enhancing citizen behavior. Striking a balance between complexity and computational feasibility is essential. Our research question is, how can we make agents behave more realistically? We assumed that happiness is a motivating factor for the mobility. Insights from a...

81. Akidu - Automatische, durch KI-integrierte Disposition für Universalterminals (129)

Poster

One aim of the Akidu project is to automatically determine the position of goods at the Unikai port terminal in Hamburg. Goods such as unit loads, cranes or other types of goods with typically abnormal dimensions might have been moved without updating their position in the terminal system in order to create free space for upcoming goods. If goods are equipped with markers such as Datamatrix...

126. Anomaly detection for discovery in particle physics (133)

Manuel Sommerhalder (Universität Hamburg, Institut für Experimentalphysik)

Poster

In their quest to understand the fundamental law governing both the cosmos and interactions between particles, physicists are searching for small deviations from established theories in massive amounts of experimental data. Searches targeting specific signatures have yielded no positive results since the discovery of the Higgs boson, raising the question of whether the right signatures have...

105. ARDIAS: AI-Enhanced Research Discovery and Advisory System (147)

Poster

ARDIAS is a web-based application that aims to provide researchers with a suite of discovery, collaboration, and recommendation tools. ARDIAS allows searching for authors and articles by name and gaining insights into the research topics of a particular researcher. It also presents a conversational interface where users can chat about a particular author or paper. Lastly, it implements a...

117. Assessment of antibiotic resistance from microbial genomic data (22)

Poster

Increasing antibiotic resistance of disease-causing microbes poses a major public health problem. Resistance is caused by evolutionary changes in the microbial genome that inactivate the antibiotic’s molecular mechanism of action. For diseases such as tuberculosis, single genetic variants that confer resistance are known and considered in treatment regimes. For other pathogens, relationships...

72. Augmenting Social Media Actor Lists with Explorative Network Sampling: Mapping German Right-wing Counterpublics on Twitter (14)

Poster

Our research addresses the challenge of accurately classifying far-right extremist accounts on Twitter/X or comparable platforms on a large scale to provide a comprehensive view of the digital far-right ecosystem. Traditional approaches rely on precise but limited account lists, leading to incomplete insights, especially for new or less popular accounts. In response, we present a mixed method...

52. Automated symptom evaluation using sensor data from wearables in patients with Parkinson's disease utilizing explainable AI (102)

Alexander Wiederhold (Universität Hamburg)

Poster

Parkinson’s disease is a neurodegenerative disorder that affects primarily dopaminergic neurons often showing a characteristic tremor or akinesis. While the severity estimation remains a stationary UPDRS scoring, we aim for an additional sensor-driven approach to develop an automated symptom evaluation for a second opinion. Our team collected over 200 million samples of accelerometer data and...

155. BACI/4C: Next-Generation Multiphysics Simulation for Challenging Real-World Problems (29)

Poster

BACI/4C is a massively-parallel multi-physics research code to analyze and solve a plethora of physical real-world problems by means of advanced computational mechanics. BACI/4C provides simulation capabilities for a variety of physical models, including single fields such as solids and structures, fluids, scalar transport, or porous media, and multi-physics coupling and interactions between...

56. bAIome Center for Biomedical AI- Towards Intelligent Health (67)

Poster

bAIome is the center for biomedical AI at University Medical Center Hamburg-Eppendorf (UKE). It consists of faculty members and staff from different institutes and departments within UKE engaged in research and education in AI relevant to biomedicine. bAIome promotes basic and applied AI research that has the possibility to translate into innovative solutions that will integrate into clinical practice.

14. Beta maṣāḥǝft: Die webbasierte Kollaborationsplattform für Handschriftenforschung (75)

Poster

Das Langzeitvorhaben der Akademie der Wissenschaften in Hamburg „Die Schriftkultur des christlichen Äthiopiens und Eritreas“ (2016-2040) entwickelt eine virtuelle Forschungsumgebung, in der detaillierte Beschreibungen von Handschriften mit Informationen zu Literaturwerken, Personen (Schreibern, Besitzern und Autoren) und Orten verknüpft werden (https://betamasaheft.eu/). Die computergestützte...

114. Biomolecular Data Science in Pneumology (23)

Poster

The Data Science group at Research Center Borstel, Leibniz Lung Center performs cutting-edge research on data generated by novel biomolecular technologies to address questions in pneumology. Towards this, we use computational methods on biological high-throughput data, so called OMICS data, to gain insights into molecular relationships underlying phenotypic traits and diseases. Focus thereby...

96. BORGES: A database for the storage, search and retrieval of ICON data (49)

Jairo Alonso Segura Bermudez (Max-Planck-Institut für Meteorologie)

Poster

BORGES tackles the handling of the ever-growing climate model data produced. It is a semantic database for the storage, search and retrieval of model data. It serves mainly demands of the up-to-date, competitive Earth system model ICON, developed and used at the MPI-M, Hamburg. BORGES excels for both large ensembles of model experiments as well as very high-res, large volume, „storm-resolving“...

32. Calculating More Property Distributions of Chemical Fragment Spaces (28)

Poster

In the early stages of drug discovery projects, it is a common task to search digital molecule collections to find promising lead structures that serve as a starting point for developing new drugs. Recently developed combinatorial compound catalogs are orders of magnitude larger than traditional databases, promising higher potential for finding relevant lead structures. Many cheminformatics...

7. Causal AI mit DoubleML (37)

Poster

Während KI aktuell hauptsächlich für Vorhersagen verwendet werden, die auf Korrelationen basieren, sind viele Fragestellungen in der Industrie und Forschung kausaler Natur. Das neue Gebiet der Causal AI / Causal ML kombiniert Methoden der kausalen Inferenz mit Methoden des künstlichen Intelligenz. Ein Ansatz dafür ist das sogennante Double Machine Learning. Es soll dieser Ansatz methodisch...

65. Climate Data Operators - CDO (58)

Poster

The Climate Data Operators (CDO) software is a collection of many operators for processing of climate and weather model data. It includes simple statistical and arithmetic, data selection, interpolation functions and many other.
The extremely large amount of data produced by high-resolution climate simulations make such problems essentially I/O bound. CDO is optimised to cope with...

45. Combining numerical methods and machine learning for physical and engineering sciences (69)

Poster

Many areas in the physical or engineering sciences rely on computational models to some extent. These models can be based on fundamental physics processe, typically leading to a set of differential equations. Alternatively, machine learning techniques can be used to infer input-output relations out of very large sets of data. Both approaches come with different strengths and weaknesses but...

68. Comparative Annotation to Explore and Explain Text Similarities (CompAnno) (122)

Poster

Das DFG-Projekt CompAnno entwickelt einen vergleichenden Annotationsworkflow zur computergestützten Detektion und Klassifizierung von literarischen Textähnlichkeiten am Beispiel von Figureneigenschaften als einer Kategorie, die sowohl für die Gestaltung literarischer Texte im Allgemeinen als auch für die Interpretation intertextueller Beziehungen zentral ist. Im Gegensatz zu etablierten...

16. Computational Analysis of Handwriting Style (9)

Poster

Images of digitised historical written artefacts contain much more information than the mere textual content found in their transcriptions, such as the texture of writing support, the general visual layout, and the non-textual visual elements. Among these aspects, the style of handwriting itself can tell us about the scribe, the approximate place of origin, and perhaps even help date the...

46. Computational aspects of the active self (80)

Poster

We, as humans, routinely talk about ourselves, but what is this “self”, how does it arise, and what influences it? What are the underlying mechanisms that help humans perceive themselves and act in the world? Recent work studies the so-called minimal self and shows how the sense of body-ownership, agency and control contribute to it. In our research, as part of the DFG SPP “the active self”,...

144. Computational Imaging@DESY (55)

Poster

We will present the work and expertise of the newly established Computational Imaging Group at DESY. Being founded in mathematics, we provide expertise in inverse problems as well as theoretical foundations of machine learning methods.
We give an overview of mathematical methods in imaging and applications thereof, in particular algorithms for reducing computational cost in large scale...

73. Conceptual foundation of machine learning in Engineering using example of tailored laser pulse forming and Optical Mirror Design (61)

Poster

Next generation plasma accelerators require ultrashort laser pulses with tailored pulse characteristics, provided at high-peak and average power.
Advancing the recent method of versatile tuning of wavelength of Ultrashort pulses further promises to open entirely new parameter regimes in particular for high-power ultrafast lasers. The approach relies on nonlinear light-matter interaction of...

20. CRIS at UHH: digital, structured, sustainable (7)

Poster

The current research information system (CRIS) of the UHH is the digital system where metadata of research activities is collected and linked in a structured way. The combined metadata in this one system creates a digital network of the actual research landscape at UHH. This network provides a fundamental base and an important tool for improving the findability, management and evaluation of...

93. CSL Poster Overview (47)

Poster

This poster shall provide an overview of the history and existing projects within our research institute at HafenCity University Hamburg. Our research projects employ innovative tools and digital city models to visualize and simulate data-driven urban developments, collaborating with a diverse range of stakeholders, including academia, policymakers, civil society, business, non-governmental...

139. D-WISE - Digitale Wissenssoziologische Diskursanalyse (150)

Poster

The D-WISE Tool Suite (DWTS) is a web-based working environment for digital qualitative discourse analysis in the Digital Humanities (DH). It addresses the limitations of current DH tools induced by the ever-increasing amount of heterogeneous, unstructured, and multi-modal data in which the discourses of contemporary societies are encoded. To provide meaningful insights from such data, the...

49. DaFne - Platform Data Fusion Generator (130)

Poster

Der Generierung synthetischer Daten kommt in der Entwicklung von KI-Methoden dort eine große Bedeutung zu, wo Trainingsdaten für diese Methoden nicht ausreichend oder in passender Form verfügbar sind. Die praktische Nutzbarkeit bisher existierender Methoden zur Datengenerierung ist oft eingeschränkt, da sie auf spezifische Einsatzgebiete zugeschnitten sind, ihre Anwendung erhebliche Expertise...

51. DASHH: Data Science in Hamburg – Helmholtz Graduate School for the Structure of Matter -> Educating the Next Generation of Data Scientists (27)

Poster

The new generation of analytical instruments produces enormous amounts of data to be analyzed and reconstructed. This development demands innovative methods at the interface of natural sciences and applied mathematics/computer science and a new generation of data scientists well-trained for interdisciplinary research.
At DASHH (https://www.dashh.org), doctoral researchers are supervised by...

95. Data analysis at the European XFEL (48)

Poster

Our poster presents an overview of the activities within the Data Analysis group at European XFEL. These encompass the development of diverse software tools designed to support the steering and interpretation of experimental data during beamtimes, as well as the post-experiment explanation of their outcomes.We introduce [DAMNIT][1], a tool engineered to automatically generate experiment...

27. Data-based tool try out in sheet metal forming (141)

Herr Michael Hohmann (HSU/UniBw Hamburg)

Poster

In sheet metal forming, the development cost and complexity of the tool try out is a major challenge. Data-based tool try out focuses on two main areas: the interpretation of die spotting images as indicators of tool quality, and the generation of tool designs based on the parts to be formed. In addition to the part geometry, various physical conditions are taken into account. Limited data...

163. DDLitLab (71)

Poster

Das von der Stiftung Innovationen in der Hochschullehre geförderte Projekt „Digital and Data Literacy in Teaching Lab“ (DDLitLab) fördert digitale Lehrinnovationen mit Bezug zur „Data Literacy Education“ durch Lehrprojekte an allen Fakultäten und im fakultätsübergreifenden Studium Generale. DDLitLab verfolgt zwei Ziele:
1. Die Weiterentwicklung der „Digital University Teaching Literacy“...

88. Decentralized Batch Effect Correction in Multi-Center Biomedical Studies: Introducing FedComBat, a Federated Learning Approach (96)

Jan Baumbach

Poster

Batch effect correction is a pivotal challenge in biological data analysis, affecting fields such as genomics, transcriptomics, and proteomics. Traditional correction methods like limma::removeBatchEffect and ComBat require centralization of data from various institutions, which can be problematic due to privacy and data governance concerns. Our innovative approach, named FedComBat,...

140. Decentralized Data Mining (44)

Poster

Obwohl die populären Social Media Plattformen sich teils stark unterscheiden, haben sie eine Eigenschaft gemeinsam: Sie sind zentral organisiert. Im Gegensatz zu diesem zentralisierten Modell steht das dezentralisierte Web (DW). Angebote des DW sollen sich jedoch nicht nur auf technischer Ebene von den etablierten Social Media Plattformen unterscheiden, sondern auch eine Alternative zur...

74. Detection of alternative splicing: deep sequencing or deep learning (94)

Poster

Alternative splicing enables the expression of a variety of isoforms coding for functionally diverse proteins from a single gene. RNA sequencing (RNAseq) has become the state-of-the-art tool for profiling the transcriptome, but still reliable detection of alternative splicing events in RNAseq from virus-infected cells with low number of reads is challenging. Few computational tools, such as ...

31. Diffusion Models for Audio-Visual Speech Enhancement (119)

Poster

This poster showcases a selection of our work on diffusion models for speech enhancement. While diffusion models have proven successful in natural image generation, we adopt them for speech enhancement by introducing a task-adopted diffusion process in the complex short-time Fourier domain. Our results show competitive performance compared to strong predictive methods, while generalization is...

103. Digital Data Handling at UWA (10)

Poster

Around 150 researcher, 40 academic disciplines, 60 research projects, and 11 research fields exist at the Cluster of Excellence UWA (Understanding Written Artefact). The size of the research field shows, it is a challenge to reconcile the many researchers with their different needs. The Research Field Data Linking has taken on the task of establishing research data management in the humanities...

25. Digital Scholarship Services at State- and University Library Hamburg (77)

Poster

The Department of Digital Scholarship Services of the State and University Library Hamburg is a partner for the conception, planning and implementation of projects in the field of digital humanities. We offer advice and support for project proposals and research projects, and act as a communication interface between the library infrastructure and Hamburg's academic institutions. We work...

87. Digitisation Services @ SUB HH (63)

Poster

Over the past 10 years, the Hamburg State and University Library has built up a comprehensive technical and organisational infrastructure for the digitisation of 2D objects for digital work with texts and images of a wide variety of materials (newspapers, journals, manuscripts, estate material, old prints, maps, copperplate engravings, etc.). We also support other institutions in Hamburg in...

127. Does Journalism Reflect Polarized Online Networks? A Comparative Analysis of the Discourse on the "Last Generation" Climate Protest Movement in (Social) Media (72)

Poster

While science emphasizes the urgency of addressing the climate crisis, political actions to combat climate change lag. Numerous climate protest movements have emerged in recent years. However, "The Last Generation," stood out in recent German media coverage.

Our research focuses on the discourse around this climate protest group, which gained significant attention due to their disruptive,...

39. DRaCOoN: Differential Regulation and CO-expression Networks (101)

Poster

To address challenges in identifying crucial regulatory elements, we introduce DRaCOoN, a data-driven method for differential co-expression and regulatory networks between unique conditions. DRaCOoN uses established metrics to better handle large datasets and offers algorithmic and benchmarking strategies for accuracy and relevance. The method employs permutation tests and a background model...

78. Ein Intelligentes Fenster zum Hamburger Hafen: Entwicklung einer Augmented Reality Fensterscheibe für Schiffe und Infotainment-Installationen (5)

Poster

Wir haben eine Augmented Reality Fensterscheibe entwickelt, die als Visualisierungassistenz auf Schiffsbrücken oder zu Infotainment-Zwecken als Outdoor-Ausstellungsstück eingesetzt werden kann. Die Umsetzung umfasst AIS-Datenempfang, Tiefenkamera zur Betrachtererfassung und ein T-OLED-Display. Durch die Anzeige von perspektivkorrekten Inhalten im Hintergrund auf einem transparenten Display im...

143. Elucidation of epigenetic mechanisms underlying normal and disease biology (2)

Poster

Epigenetics is defined as the study heritable alterations in gene function that occur without underlying changes in DNA sequence. Biologically, aging is linked with a gradual increase in molecular and cellular damage eventually leading to a decline in physiological reserves and an increased risk of developing diseases. It is now well established that a vast number of epigenetic changes,...

129. Enabling Secure Communication for Cyber-Physical Processes in Critical Infrastructures (41)

Poster

Machine-to-machine communication over wireless networks is increasingly adopted to improve service and maintenance processes at airports, ports, and manufacturing plants. This brings with it the challenge of how to bootstrap a secure communication channel between the machines involved. Building on the idea of secure device pairing we research novel schemes for key establishment that exploit...

160. Entwicklung eines computergestützten Tools zur Bereitstellung von Informationen, Struktur und Entscheidungshilfen für das hausärztliche Management von Multimorbidität (gp-multitool.de) (33)

Poster

Hintergrund
Um das hausärztliche Management älterer Patienten mit Multimorbidität zu unterstützen, wurde ein digitales Tool entwickelt, das ermöglicht, behandlungsrelevante Informationen zu erfassen und zu dokumentieren.
Methoden
Funktionalitäten und Anwenderfreundlichkeit des Tools wurden in Fokusgruppen mit Hausärzten und Patienten diskutiert. Das Tool wurde als Webapplikation...

36. ERIS: Hamburg Information System on the Representation of Greek and Roman Violence (98)

Poster

Das MyCoRe-basierte Informationssystem Eris, welches seit 2012 unter der Leitung von Prof. Dr. Werner Rieß von Mitarbeiter*innen aus dem Arbeitsbereich Alte Geschichte an der Universität Hamburg aufgebaut wird, ermöglicht zum ersten Mal eine inhaltliche und multidimensionale Erschließung aller historiographischen und biographischen Texte zur antiken Gewalt.

28. EvalSpek-ML (140)

Poster

Compositional analysis identifies the elemental makeup of materials, with applications including tracing the geological origins of archaeological artifacts. Spectra, encompassing for example 16,384 distinct energy channels, provide a detailed elemental composition when analyzed by experts. This process is complex and time consuming, therefore EvalSpek-ML aims to use machine learning algorithms...

44. ExaOcean: Improving Performance of the ICON-O Oceanmodel on heterogeneous Exascale-Supercomputers with Machine Learning (68)

Poster

Ocean models are a key component of every weather or climate model. Despite the computing power of modern supercomputers, however, important dynamical features can so only be resolved in simulations over a few weeks.

ExaOcean will deliver modern mathematical algorithms to achieve better parallel scaling and faster runtimes in highly resolved simulations on new supercomputers. We will...

159. Exploratory Clinical-Lab Data Analysis (32)

Poster

Laboratory test results play a significant role in clinical decisions for individual patients. Analysing these results over large populations and extended periods could offer additional insights. The UKE has collected lab results for 24 years, and hundreds of thousands of patients; however, this wealth of data has yet to be explored. Decomposing the factors that contribute to the observed...

148. Exploring supervised machine learning algorithms for predicting pitting corrosion damage of AA1050 exposed to ethanol-containing fuels (107)

Poster

Aluminum alloys are crucial in car manufacturing, but the rise of bioethanol fuels has raised concerns about alcoholate corrosion. Gazenbiller et al. have already explored temperature-induced alcoholate pitting corrosion in AA1050 aluminum exposed to anhydrous ethanol.1 To extend this study, a specially constructed reactor is used, which allows for in situ tracking of chemical corrosion damage...

92. Exploring Ultra-Large Chemical Spaces With Genetic Algorithms (26)

Poster

A major challenge in modern drug design is the vast number of possible molecules that have to be navigated to find a few molecules of interest for a particular project.

Robust search heuristics like genetic algorithms can elevate established methods in the realm of cheminformatics to find this figurative needle in a haystack. Our approach, Galileo, finds promising hit compounds in...

133. Extending the Public DGS Corpus in Size and Depth (124)

Poster

In 2018 the DGS-Korpus project published the first full release of the Public DGS Corpus with new data formats, corpus annotation conventions, and OpenPose pose information all transcripts. The community and research portal websites of the corpus also received upgrades, including persistent identifiers, archival copies of previous releases and improvements to their usability on mobile devices....

23. Facilitating FAIR Data Practice: The Research Data Repository for the Universität Hamburg (6)

Poster

Following the FAIR principles helps make data findable by humans and machines, interoperable through standard formats and metadata, accessible through open access when possible, and reusable through clear licensing and richness of metadata, supporting improved understanding, reuse, and reproducibility of scientific data.
While following these principles requires the willingness to embrace...

90. FAIR-Based Strategy for Data Management and Handling in Large Climate Modeling Projects: Experiences from PalMod-II (99)

Poster

Here we present the research data management (RDM) approach developed and employed for project PalMod-II, making large scale climate model data available for reuse by the global paleo-climate science community in-line with the FAIR (Findable, Accessible, Inter-operable, Reusable) data principles. The compilation and maintenance of a project-wide data management plan (DMP) was prepared and...

131. Fast machine learning in particle physics (134)

Poster

Particle physics studies the nature of elementary particles. There are strong hints that the current theories are not complete yet and that so far undiscovered particles exist. They could be produced in proton collisions at very high energy. To allow significant production of new particles in reasonable time, collisions are performed at 40MHz. Only a small fraction of the data is stored,...

5. Feature Reduction Improves Drug Response Prediction (111)

Poster

Machine learning approaches play an important role in precision medicine by predicting drug responses based on the molecular profiles of patients. However, the high dimensionality of molecular profiles requires feature reduction. To assess the effectiveness of different feature reduction strategies for drug response prediction (DRP) on cancer, we evaluated four feature reduction methods on...

13. Federated Machine Learning for Big Data (74)

Poster

On this poster, we are going to present the research of the Institute for Data Engineering at TU Hamburg on the processing of very large data streams as well Federated Learning (FL). Federated Learning (FL) offers an alternative approach to centralized Machine Learning, by distributing the model generation across different entities. Thus, learning can be conducted close to the data sources,...

9. Few-shot learning for automated content analysis (FLACA) in the German media debate on arms deliveries to Ukraine (12)

Poster

The use of pre-trained language models based on transformer neural networks has significantly advanced the field of NLP and offers considerable potential for improving automatic content analysis, e.g., in communication science, where their widespread adoption is still limited. In our poster, we highlight challenges and promises by employing transformer models combined with parameter-efficient...

115. FLIGHT: Federated LearnIng Guided digital Health (157)

Poster

Around 60,000 men are annually diagnosed with prostate cancer in Germany, which makes it the second-most frequent cancer. We previously developed the state-of-the-art analysis method eCaReNet (explainable Cancer Relapse prediction Network) for survival prediction of prostate cancer patients based on tissue microarray (TMA) data. To build a more robust and accurate model, data from multiple...

119. Flimma: a federated and privacy-aware tool for differential gene expression analysis (83)

Poster

Aggregating transcriptomics data across hospitals can increase sensitivity and robustness of differen-
tial expression analyses, yielding deeper clinical insights. As data exchange is often restricted by pri-
vacy legislation, meta-analyses are frequently employed to pool local results. However, the accuracy
might drop if class labels are inhomogeneously distributed among cohorts. Flimma...

132. From Super-Resolution to Downscaling - An Image-Inpainting Deep Neural Network for High Resolution Weather and Climate Models (70)

Poster

High resolution in weather and climate was always a common and ongoing goal of the community. In this regards, machine learning techniques accompanied numerical and statistical methods in recent years. Here we demonstrate that artificial intelligence can skilfully downscale low resolution climate model data when combined with numerical climate model data. We show that recently developed image...

91. FUNDus meets HCDS - Novel avenues to explore scientific collections (151)

Poster

Die Sichtbarmachung von derzeit insgesamt 31745 Objekten aus 59 Sammlungen mit 32404 Digitalisaten wird durch die enge Zusammenarbeit des Zentrums für nachhaltiges Forschungsdatenmanagement (ZFDM) mit der Zentralstelle für wissenschaftliche Sammlungen und dem LIB (ehemals Centrum für Naturkunde) ermöglicht.

In Kooperation mit dem House of Computing and Data Science (HCDS) arbeiten wir...

121. Generative Machine Learning for Particle Physics (132)

Thorsten Buss (Institut für Experimentalphysik, Universität Hamburg)

Poster

To analyze events in particle colliders, a comparable amount of events has to bee simulate as events are recorded. The Monte Carlo simulation of the detector needs most of the computing resources. However, the required resources will exceed the available ones soon.
To tackle this problem, we are developing several generative machine learning models in our research group at the Institute for...

83. Genome-based prediction of autoimmune diseases using relation neural networks (24)

Poster

The development of autoimmune diseases arises from a complex interplay of genetic predisposition and environmental influences. Deep learning based approaches could boost the predictive performance by capturing non-linear relationships between genetic variants and the phenotype. In this project, I present and evaluate a deep neural network architecture based on Relation Neural Networks, a...

102. German Large Language Legal Model (GLLLM) (146)

Chris Biemann (House of Computing and Data Science), Dr. Martin Semmann (House of Computing and Data Science)

Poster

Our project's primary focus is to systematically collect legal documents in German from European Union administrative tiers. This effort enables seamless access to German legislation for researchers, lawyers, and citizens alike. Using intuitive Python scripts, we've automated data collection from publicly accessible websites, streamlining the process for users. Furthermore, we employ...

77. Group dynamics in the Metaverse: A conceptual model and first empirical insights (155)

Poster

The metaverse offers new technological affordances to conduct 3D immersive meetings with head-mounted displays that can enrich virtual teamwork. We present a conceptual model of effective group interactions in the metaverse, along with a pilot study that begins to explore relevant design factors, attendee experiences, and behavioral group dynamics. Our findings show that participants intensely...

151. Helmholtz Imaging@DESY (54)

Poster

This poster will provide an overview of the Helmholtz Imaging Platform at DESY and its major activities.
Helmholtz Imaging involves both service for practitioners on different tasks in digital imaging as well as cutting edge research. The group at DESY focuses at the early part of imaging, i.e. the image formation part and related tasks such as image reconstruction, denoising, motion...

147. Herbie: ELN and research database from Hereon (45)

Anke-Lisa Höhme (Helmholtz-Zentrum Hereon)

Poster

Herbie is an Electronic Lab Notebook (ELN) developed at Helmholtz-Zentrum Hereon. Highly structured, yet flexible protocols for parameters and results of individual manufacturing and characterization processes are described with machine-readable ontologies in RDF using OWL and SHACL. By interlinking the protocols and putting them into context with involved staff, samples, equipment, projects,...

156. High-throughput processing for diffraction data (30)

Poster

Synchrotron X-ray diffraction (XRD) experiments are a versatile tool in understanding material properties and processes, for example light-weight materials or additive manufacturing and for generating data to create digital twin models.
The analysis of XRD data is often still the domain of experts because software tools were designed for flexibility with numerous parameters which makes them...

161. House of Computing and Data Science (148)

Poster

As a central institution of the University of Hamburg, the House of Computing and Data Science (HCDS) supports interdisciplinary research and application of innovative digital methods in close cooperation with its partners from science and industry in the Hamburg metropolitan region. It coordinates and supports the implementation of the digital strategy in research at the University of...

38. hpc.bw – Competence Platform for Software Efficiency and Supercomputing (dtec.bw) (139)

Poster

The project „hpc.bw – Competence Platform for Software Efficiency and Supercomputing“ (dtec.bw) aims to strengthen innovative cross-location research in the field of high performance computing (HPC) and to romote the transfer of relevant expertise to a wide range of disciplines. The established HPC Competence Platform (HPCCP) and container-based HPC computing center (CBRZ) are leveraged by...

58. Hyperbolic Learning on Omics Data (106)

Poster

Many objects in biology are connected by hierarchical relationships. To clarify how snippets of data are associated, we apply embeddings, that is, mapping of multidimensional objects into space so that similar objects are positioned at close points. Analyzing such data with tools operating in Euclidean spaces is problematic as the tools may not account for the underlying data hierarchy. We...

79. Identification of differentially expressed biclusters for unsupervised patient stratification (95)

Poster

Unexplored molecular heterogeneity of human diseases causes treatment inefficacy and hinders the investigation of causative disease mechanisms. Since the number and frequencies of disease subtypes are usually unknown, unsupervised methods are applied to omics data to identify patients subgroups with similar molecular profiles. Here, we present UnPaSt, a novel biclustering algorithm for...

112. In the Pursuit of Eco-Friendly Wind Energy – Unmanned Aerial Systems and Computational Strategies (38)

Poster

The project Drones4Bats, funded by the BMWK, employs computational methodologies with autonomous unmanned aerial systems (UAS) to enhance bat-friendly wind turbine systems (WTS) operation and increase wind energy output. It involves developing and comparing various UAS technologies, including Multicopters, specialized bat-friendly UAS, and lighter-than-air UAS. Key computational methods...

48. In-the-wild Speech Emotion Conversion Using Disentangled Self-Supervised Representations and Neural Vocoder-based Resynthesis (121)

Poster

Speech emotion conversion aims to convert the expressed emotion of a spoken utterance to a target emotion while preserving the lexical information and the speaker's identity. In the context of human-machine interaction systems (e.g., social robots), to improve the naturalness of machine communication, the generation of emotionally expressive speech is required. In this work, we introduce a...

62. INEL – Grammar, Corpora and Language Technology for Indigenous Northern Eurasian Languages (57)

Poster

Ever since 2016, the spanning 18 years long-term project INEL has been generating deeply annotated language corpora and accompanying digital resources using existing and acquiring new language material from a number of heavily endangered languages of the Northern Eurasian Area. The core aim is not only to make these resources sustainably available after their publication but also provide...

15. Inferring Tissue-Specific Alternative Splicing from Blood Samples (115)

Poster

This study explores an innovative, less invasive approach to diagnosing complex diseases like dilated cardiomyopathy (DCM) by predicting target tissue expression profiles using gene expression and alternative splicing profiles from blood samples. Machine learning approaches, specifically dimensionality reduction and linear regression, are utilized. Preliminary results indicate that while this...

113. Inflation Narratives from a Machine Learning Perspective (135)

Cedric Möller (Universität Hamburg), Junbo Huang (Universität Hamburg)

Poster

Inflation narratives explain inflation changes and affect expectations. Manually identifying them is cumbersome, prompting the need for scalable algorithms. Narratives comprise events, causal relations, and arguments, represented as graphs with event and argument nodes. Causal relations indicate cause-and-effect relationships between events using directed edges. Our main objective is to...

42. Inflation Narratives from an Economics Perspective (136)

Poster

Previous research in the research group has focused on two main areas: First, building on the
work of Andre et al. (2023), Eliaz & Spiegler (2020) and others, a survey has been conducted
and narratives have been represented in the form of "directed acyclic graphs" (DAGs, Pearl,
2009). Second, based on the DAGs of Andre et al. (2023), a study of the distribution of
narratives in a large...

41. Institute for Data Science Foundations (79)

Poster

This poster outlines the research and teaching profile of the Institute for Data Science Foundations at the Hamburg University of Technology.

12. Integrated Climate Data Center (88)

Poster

The Integrated Climate Data Center (ICDC) is an established service component of the Center for Earth System Science and Sustainability (CEN) of the University of Hamburg (http://www.cen.uni-hamburg.de/icdc). Our main mandate is to provide easy-to-use observational (in situ, remote sensing) data products describing the Earth's climate system, to consult scientists across all career levels in...

146. InteGreatDrones: New Perspectives on Ports Through Drones and AI (154)

Poster

The rise in production of cost-efficient unmanned aerial vehicles ("UAVs" or "drones") opens up new possibilities for process optimizations in industrial settings. The InteGreatDrones project focuses on the use of drone swarms for data collection in dynamic inland terminals, addressing challenges of changing cargo types and stakeholder interactions. The project's goal is to employ autonomous...

69. Interdisciplinary and International Collaborations: The MARS Research Group at Hamburg University of Applied Sciences (HAW) (60)

Poster

The MARS Research Group at Hamburg University of Applied Sciences fosters interdisciplinary and international collaboration. Our projects address diverse demands, from city planning and logistics to global change mitigation. We have developed an extensive framework for large-scale agent-based simulations, extensively used in both our research and global university courses.
Our current work...

54. Introduction of the research group "Chemometrics of complex material systems" (19)

Poster

In Stephan Seifert's research group "Chemometrics of Complex Material Systems" at the Hamburg School of Food Science, analytical data corresponding to fingerprints of biological samples are analysed. These fingerprints come, for example, from food or written artefacts, which are classified using machine learning methods based on various properties, such as their geographical origin. The...

164. KI für die Internationale Raumstation (128)

Poster

In der technologisch komplexen Umgebung der Internationalen Raumstation (ISS) ist die schnelle und akkurate Identifikation von Fehlerursachen sowie deren Behebung, insbesondere bei Störungen im Lebenserhaltungssystem, kritisch. Aktuell stellen die Analyse von Daten aus mehr als 20.000 Sensoren und das Verständnis der komplexen Wirkzusammenhänge in der Station eine bedeutende Herausforderung...

157. KI4D4E: Artificial Intelligence for Synchrotron-radiation 4D Tomography Data (31)

Poster

The Helmholtz-Zentrum Hereon is operating imaging beamlines for X-ray tomography (P05 IBL, P07 HEMS) for academic and industrial users at the synchrotron radiation source PETRA III at DESY in Hamburg, Germany. The high X-ray flux density and coherence of synchrotron radiation enable high-resolution in situ/operando/vivo tomography experiments. Here, large amounts of 4D data are collected from...

141. Legal Hypergraphs (91)

Poster

Network analysis has been widely adopted to investigate law as a complex system. However, the utility of dynamic higher-order networks in the legal domain has remained largely unexplored. Setting out to change this, we introduce temporal hypergraphs as a powerful tool for studying legal network data. Temporal hypergraphs generalize static graphs by (i) allowing any number of nodes to...

21. Leveraging Jupyter on Maxwell HPC: joyful, visual and green computing (76)

Poster

Jupyter notebooks are great tools to mitigate the complexities of (heterogeneous) HPC systems - like the Maxwell cluster at DESY which serves the computational needs of all user facilities on campus, as well as a wide variety of applications ranging from plasma accelerators to quantum chemistry. We aim to expand the Jupyter ecosystem utilizing python application frameworks to provide...

122. Leveraging Semantic Information for Efficient Self-Supervised Emotion Recognition with Audio-Textual Distilled Models (110)

Poster

In large part due to their implicit semantic modeling, self-supervised learning (SSL) methods have significantly increased valence recognition performance in speech emotion recognition (SER) systems. Yet, their large size may often hinder implementation in applications such as virtual assistants and digital customer service. In this work, we analyze the relevance for SER of each of HuBERT's...

104. Machine Learning based optimization of transverse beam emittance in Free Electron Lasers (81)

Poster

In the context of Free Electron Lasers like the European XFEL at DESY Hamburg, it is essential to optimize the transverse emittance of a charged particle beam. Within the project OPAL-FEL we therefore aim to implement a Machine Learning based online optimal control process to minimize emittance.
The process will consist of a forward prediction model and an inverse feedback model.
Their...

153. Machine Learning for the Automated Selection and Reconstruction of Multi-Modal Nanotomography Data of Bone-Implant Interfaces (87)

Poster

Synchrotron radiation-based imaging and structural characterization provides unique opportunities to study hierarchical materials systems, such as bone. Correlative 3D imaging enables linking cellular connectivity and bone mineralization via transmission X-ray microscopy, X-ray diffraction and X-ray fluorescence. Generated data is terabyte-scale. We have acquired multi-modal datasets of bone...

33. Machine Learning Group at Universität Hamburg (144)

Poster

The Machine Learning Group at Universität Hamburg is at the forefront of research in the field of efficient machine learning. Our specialized focus lies in developing cutting-edge optimization methods that enhance the training of machine learning models. Leveraging innovative techniques and algorithms, our group strives to make machine learning more efficient, scalable, and sustainable. We are...

138. Making Academies' long-term research projects ready for the NFDI (43)

Poster

The poster presentation will provide an overview of the digital infrastructure and strategy developed at the Academy of Science and Humanities in Hamburg as part of the activities in the framework of the NFDI cluster Text+. To this end, a brief presentation of the six projects funded by the German Academies' long-term program in Hamburg as well as their respective research goals and...

124. marFM®, A Multilingual Automatic Speech Recognizer (ASR) for VHF-Radio Communication (40)

Frau Sophia Lichtenberg (Fraunhofer CML)

Poster

Radio communication is a central task and safety-critical for international shipping. Despite being based on resilient and robust technology, the quality of transmitted audio signals varies depending on antenna height and prevailing weather conditions. In addition, shipping is characterized by a high degree of crew multi-nationality, so that communications over VHF radio contain a variety of...

118. Massively Parallel Hybrid Molecular-Continuum Simulations (39)

Poster

Micro- and nanofluid flow simulations require considering some effects at the molecular scale. Molecular-continuum coupled flow simulations can perform computationally intensive molecular dynamics (MD) simulations in localized regions of a geometry under consideration, and employ classical, computationally cheaper computational fluid dynamics (CFD) solvers for the remaining larger...

63. Mikropolis - Orientierung in der digitalen Transformation (153)

Poster

Mikropolis versteht sich als Digital Literacy-Projekt mit der Zielsetzung, Studierenden Orientierung und Urteilsfähigkeit über den Prozess der digitalen Transformation zu vermitteln und dieses Lernziel mit der nachhaltigen Gestaltung der Digitalisierung (sustainable development Goals/ESG-Kriterien) zu verbinden.
Zielsetzung ist, der Transfer der Plattform MikroPolis.org sowie die inhaltliche,...

98. ML based process monitoring of semi-automatic drilling processes (51)

Poster

Since one third of rivet holes during aircraft assembly are produced with semi-automatic drilling units, in this work reliable and efficient methods for process state prediction using Machine Learning (ML) classification methods were developed for this application. Process states were holistically varied in the experiments, gathering motor current and machine vibration data. These data were...

22. Morphology-based Molecular Classification of Spinal Cord Ependymomas Using Deep Neural Networks (142)

Poster

Introduction

Ependymomas are tumors of the central nervous system and DNA methylation profiling is used to distinguish them into 10 molecular types. However, traditional assessment by neuropathologists of tumor tissue stained with hematoxylin and eosin (H&E) does often not match to the correct molecular group. Focusing on the molecular ependymoma types myxopapillary (MPE) and spinal...

66. Motion Editing tool for Reproducing Grammatical Elements of Japanese Sign Language Avatar Animation (59)

Hiroyuki Kaneko (NHK Science & Technology Research Laboratories), Masanori Sano (NHK Science & Technology Research Laboratories), Naoki Nakatani (NHK Science & Technology Research Laboratories), Taro Miyazaki (NHK Science & Technology Research Laboratories)

Poster

Avatar animation is one of the ways to produce sign language content. Using motion capture data allows for fluent avatar movements closer to human signer. However, the generated animation usually cannot adjust grammatical features such as intonation, signing space, and facial expression to match a specific context. This is because all motion data is in a fixed form at the time it is...

109. MRI Total: Towards an automated, standardized, reproducible, version-controlled, transparent, FAIR, quality-assured, privacy-compliant processing pipeline for MRI data (53)

Poster

Neuroscientists at University of Hamburg (UHH) collect thousands of hours of human brain imaging data per year. However, this data treasure is not systematically standardized, stored, or controlled for quality. We propose an institution-wide processing pipeline for human brain imaging data that connects state-of-the-art open-source neuroinformatics software tools and leverages UHH...

136. Multilingual Racial Hate Speech Detection Using Transfer Learning (145)

Poster

The rise of social media eases the spread of hateful content, especially racist content with severe consequences. In this paper, we analyze the tweets written in French targeting the death of George Floyd in May 2020 as the event accelerated debates on racism globally. Using the Yandex Toloka platform, we annotate the tweets into categories as hate, offensive, or normal. Tweets that are...

40. Multimodal Modelling of Cultural Artefacts in Digital Space (120)

Poster

The poster presents the cross-disciplinary methods from artificial intelligence and humanities that we plan to apply in the project MuMokA, which deals with the multimodal modeling of cultural artefacts. We apply digital methods from the area of digital restauration, natural language processing, computer vision, and affective computing to extract modalities from the unfinished multimodal work...

108. MY DGS – ANNIS: ANNIS and the Public DGS Corpus (117)

Poster

The Public DGS Corpus is a collection of annotated German Sign Language data which is available through several different interfaces. In 2022 we published our third data portal (MY DGS – ANNIS), using the ANNIS browser-based corpus software. ANNIS is a corpus query tool for visualization and querying of multi-layer corpus data which has its own query language, AQL, and is accessed from a web...

142. Natural Language Processing in the Legal Domain (82)

Poster

We summarize the current state of the field of NLP and Law with a specific focus on recent technical and substantive developments. To support our analysis, we construct and analyze a corpus of more than six hundred NLP and Law related papers published over 10 years. Our analysis highlights several major trends. Namely, an increasing number of papers written, tasks undertaken, and languages...

8. NeDRex - an integrative and interactive network medicine platform for drug repurposing (113)

Poster

Drug discovery's prevailing paradigms grapple with efficacy challenges. Drug repurposing, with its agility and cost-efficiency, emerges as a potent alternative. Complex diseases are unraveled through disease modules, highlighting potential drug targets discernible in silico via network algorithms. However, essential data remains dispersed across various databases, underscoring a demand for...

53. On the power and limits of structure-based exon-exon interaction prediction (103)

Poster

To perform important cellular functions, Proteins physically interact with each other through protein-protein interactions (PPIs). However, PPIs are not constant but change with varying cellular conditions, e.g., those induced by infections or other diseases. These changes in PPI can be caused by alternative splicing (AS), a biological process, by which a single gene gives rise to multiple...

120. Open Access Publishing for Hamburg Research. Quality-Assured Scholarly Open Access Publication Services (89)

Poster

The digital publication of research results is an important step in the life cycle of academic research. Today, a publication needs to be as open as possible in order to unleash its full potential. In the context of the Open Access (OA) transformation institutional publishing services are emerging that enable quality-assured OA publications according to professional standards. 
We will...

17. Palimpsests inpainting with Generative AI (17)

Jost Gippert (Universität Hamburg), Dr. Mahdi Jampour (Universität Hamburg)

Poster

A palimpsest is a manuscript page that has been scraped or washed off in preparation for reuse, in the form of another document. When working with palimpsest manuscripts, even with the employment of Multispectral Imaging (MSI), the under-text often remains challenging to read. In some cases, it might not be readable at all. This difficulty arises due to various factors, including parts of the...

18. Pattern Analysis Software Tools (8)

Poster

The research of ancient written artefacts results in an ever-increasing amount of digital data in various forms, ranging from raw images of artefacts to automatically generated data from advanced acquisition techniques. The manual analysis of this data is typically time-consuming and can be subject to human error and bias. Therefore, a set of Pattern Analysis Software Tools (PAST) has been...

67. Predicting COVID-19 Vaccination Uptake from Public Discourse: A Machine Learning Approach (36)

Poster

This project explores the relationship between public discourse and COVID-19 vaccination uptake and how to use real world data to identify public opinion on COVID-19 vaccination. The analysis will apply machine learning techniques to Twitter data and will link these to data on vaccination rates. In Machine Learning the focus is mainly on providing point forecasts. But in many real-world...

134. Predicting multiple conformations of flexible proteins (85)

Poster

We developed a pipeline for predicting multiple conformations of flexible proteins. A range of conformations is first generated using the deep learning model AlphaFold2 (AF2), which are then filtered using distance constraints and solvent-accessibility data from crosslinking mass spectrometry (XL-MS), using two scoring functions we developed: the crosslink and monolink probability scores (XLP,...

150. Prediction of therapeutic targets to halt inflammation-induced neurodegeneration. (46)

Poster

Inflammation-induced neurodegeneration poses a clinical challenge in multiple sclerosis and aging. Despite intense research neuroprotective therapies are not available. Applying neuronal networks on “Omics” data holds the promise to identify novel therapies. We aim to establish a platform to find and prioritize genes that determine neurodegeneration. First, we will use generative neuronal...

137. Presenting a multi-layered manuscript of a Byzantine Greek lexicon within a responsive digital edition (42)

Poster

Our poster illustrates how the features of the Vaticanus Barberinianus graecus 70 (11th c. CE, from Southern Italy) can be reflected in a responsive digital edition currently developed within the Etymologika project. This remarkable written artefact's complexity and multi-layered nature put high demands on data modelling and GUI development.
The web interface includes a view of the...

71. Privacy-Preserving Federated Differential Protein Expression Analysis with FedProt (93)

Poster

Integrating distributed patient-derived proteomics data poses privacy concerns, risking genotype reconstruction attacks. To enable privacy-preserving analysis of distributed data, we developed FedProt, the first tool for federated differential protein expression analysis. Based on DEqMS and utilizing the hybrid methodology of federated learning and additive secret sharing from Flimma, FedProt...

125. Privacy-preserving federated prognostic model learning from multi-centric questionnaire data to predict pain changes in osteoarthritis (84)

Poster

Multi-centric patient-derived data typically is decentralized and can not be integrated into one datacenter for privacy reasons. We address this problem using federated learning to train logistic, linear, and random forest regression models in a privacy-preserving fashion across data from multiple centers. We utilize data from GLA:D® osteoarthritis registries and compare federated models...

84. Profiling of written artefacts with bioinformatic and chemometric methods (18)

Poster

Written artefacts are studied at the CSMC with the participation of scientists from a variety of disciplines using a wide range of material analysis methods. We contribute to this collaboration with various data analysis techniques, including machine learning and statistical approaches. They are aiming at the classification of the written artefacts based on different properties, e.g....

94. ProteinsPlus: On-The-Fly Structure-Based Design on the Web (25)

Poster

The ProteinsPlus web server (https://proteins.plus) [1] offers modelling support for in-depth investigation of biomolecules. Users can perform computational studies for experimental protein structures from the Protein Data Bank (PDB) [2] and predicted models from the AlphaFold Protein Structure Database [3].
The services include structure quality analyses, structure preparation, geometric...

99. PUNCH4NFDI - Facilitating FAIR data in Germany (52)

Poster

PUNCH4NFDI (the NFDI consortium of particle, astro-, astroparticle, hadron and nuclear physics) has as its central deliverable a community-overarching science data platform (SDP), in which complex workflows can be executed on digital research products (DRPs) in a transparent, automatised and FAIR (Findable, Accessible, Interoperable, Reusable) way. The SDP consists of several ingredients, not...

29. QLK – Queer-linguistisches Korpus (78)

Poster

The poster presents the project QLK – Queer Linguistic Corpus. The goal of the project is to build a queer linguistic corpus. The combination of queer linguistics, corpus linguistics, and data science represents a promising and challenging approach that is largely a desideratum for German. The project is applied and transfer-oriented and aims to build a research resource with a variety of...

80. Redirected Walks through the European XFEL (4)

Poster

At modern particle accelerators, robots are playing an increasingly important role in facility inspection and maintenance to significantly improve the working conditions of technicians and scientists. Robot teleoperation using virtual reality (VR) technology offers a faster, more accurate, and more secure way to teleoperate robots in hazardous environments. Therefore, VR interaction techniques...

154. Reproducible Knowledge Graph workflows for embedding chemical entities and associated biology of diseases: A use case in Mpox (3)

Poster

The outbreak of Mpox virus (MPXV) infection in May 2022 is declared a global health emergency by WHO. The MPXV pathophysiology and its underlying mechanisms are not yet understood. Likewise, the knowledge of biochemicals and drugs used against MPXV and their downstream effects is sparse. Using systems biology approaches, we used Knowledge Graph (KG) representations to depict chemical and...

89. RESCUE-MATE - Rethinking Innovation in Natural Disasters (152)

Poster

A consequence of drastic climate change is increased extreme weather events. One of those are storm surges. Hamburg, Germany, is well-known for its dramatic storm surges in 1962, with over 280 deaths. Since then,
the measures taken to prevent disasters and cope with the weather have matured. Nevertheless, the application of state-of-the-art information systems falls short. The project...

57. Research data management at the German Institute for Global and Area Studies (GIGA) (105)

Poster

Um einen kontinuierlichen Service im Forschungsdatenmanagement (FDM) am German Institute for Global and Area Studie (GIGA) zu gewährleisten, gibt es dort seit 2014 eine zuständige Person für diesen Themenbereich. 2017 führte das GIGA eine institutseigene Forschungsdatenleitlinie ein. Seitdem wurden und werden Kooperationen zu Infrastrukturpartnern aufgebaut und ein andauernder Kulturwandel...

111. Research in Geoinformatics and Geovisualization at HCU (159)

Poster

Research at the Lab for Geoinformatics and Geovisualization (g2lab) is about
- visualizing spatio-temporal data to enable the understanding of complex problems and decision making
- applying task-oriented approaches to generate effective and efficient visualizations
- modeling and communicating uncertainties to make decisions more certain

This poster shows selected research and PhD...

116. Scholarly Publishing in the European Research Area (ERA) - Supporting the Visibility of Hamburg Research Results (90)

Poster

Open Science is on the political and funding agenda on the national (DFG, BMBF, KMK) as well as on the international level. The European Commission funds the projects CRAFT-OA, DIAMAS, and PALOMERA in order to establish diamond open access publishing services. Networked institutional infrastructure for journals and books is being built on high quality standards and will be integrated in...

152. Selection of Effective Corrosion Inhibitors for Aluminium Alloys based on Data-Driven Techniques (108)

Poster

The replacement of highly protective but toxic hexavalent-chromium-based corrosion inhibitors with novel and safer inhibitors for aluminum (Al) alloys is urgently required. Small organic molecules have emerged as safe and potent alternatives showing promising corrosion inhibition for Al alloys [1]. Experimental techniques alone can only screen a tiny fraction of the vast chemical space of...

85. Semi-Automatic Subtitle Alignment for Sign Languages (125)

Poster

To provide subtitles for videos in sign language the translation has to be aligned with the signing.
Doing this manually is time consuming but fully automatic alignment is often not accurate enough. We therefore propose an interactive tool that uses human feedback to optimize the alignment process. For the automatic alignment we use signs detected by a sign spotter and map them to words in...

145. Shaking out Intrinsic Structural Variation in Femtosecond Protein Crystallography using Deep Learning (86)

Poster

Current femtosecond crystallography data processing routines such as CrystFEL by T. White generally assume that identical molecular structure for each of the tens of thousands of protein crystals used. This assumption is however known to be unphysical, and with the recent development of deep learning technology, we are exploring whether we can build a net capable of shaking out the subtle...

34. Sind wir interessant für KI – oder: ist KI interessant für uns? Analysen des Textkorpus „Mehrsprachigkeitsentwicklung im Zeitverlauf (MEZ)“ (65)

Poster

Die MEZ-Studie untersucht die Entwicklung mehrsprachiger Literalität bei n=2103 Schüler(inne)n der Sekundarstufe (Anfangsstichprobe) aus deutsch-russischen, deutsch-türkischen oder einsprachig deutschen Familien. Sie lernten Englisch als erste, Französisch oder Russisch als zweite Fremdsprachen. Schriftsprachliche Fähigkeiten wurden in vier Messzeitpunkten getestet. Bewertet wurden die...

101. Smart and sustainable cities? Testing a Socio-Technical Transitions Model of the Interplay of Urban Digitalization and Sustainable Development (20)

Poster

Facing challenges of climate change and urbanization cities worldwide turn to new means of digitalization aiming to increase city sustainability. However, the way urban digitalization affects sustainability is less clear. Drawing on socio-technical transitions theory we introduce a model of the dynamic interplay of digital and sustainable transformation. We aim to test that model using...

60. Smart Assembly im Labor für intelligente Leichtbauproduktion (138)

Poster

Im "Labor für intelligente Leichtbauproduktion" (LaiLa) werden in Kooperation zwischen der HSU und der CTC GmbH digitalisierte Produktionssysteme für Bauteile und Großstrukturen aus Faserverbundmaterial in der Luftfahrt vernetzt und durch künstliche Intelligenz (KI) und Maschinelles Lernen (ML) optimiert. Im Use-Case "Smart Assembly" innerhalb von LaiLa kommen intelligente Tools und...

75. SmartShip: KI für Echtzeit-Anomalieerkennung und automatischer Objekterkennung in maritimen Such- und Rettungsaktionen (137)

Alexander Windmann (Helmut-Schmidt-Universität)

Poster

In dieser Studie entwickeln wir ein KI-gestütztes System für maritime Such- und Rettungsaktionen. Das System führt in Echtzeit eine Analyse sensorischer Daten und Verhaltensanomalien von Schiffen durch und integriert automatische Objekterkennung zur Identifikation von Personen in Not. Ziel ist eine präzise Vorhersage für Wartungsarbeiten, eine genauere Zustandsüberwachung und effizientere...

10. Social Media Observatory (13)

Dr. Felix Victor Münch (Leibniz-Institut für Medienforschung | Hans-Bredow-Institut)

Poster

In unserem Poster stellen wir das Social Media Observatory (SMO) am Leibniz-Institut für Medienforschung | Hans-Bredow-Institut vor. Das SMO wird seit 2020 als Open-Science-Forschungsinfrastruktur innerhalb des Forschungsinstitut Gesellschaftlicher Zusammenhalt (FGZ) entwickelt. Es konzentriert sich auf die langfristige Beobachtung der öffentlichen Kommunikation auf ausgewählten Plattformen...

55. sPLINK: a hybrid federated tool as a robust alternative to meta- analysis in genome-wide association studies (104)

Poster

Meta-analysis has been established as an effective approach to combining summary statistics of several genome-wide association studies (GWAS). However, the accuracy of meta-analysis can be attenuated in the presence of cross-study heterogeneity. We present sPLINK, a hybrid federated and
user-friendly tool, which performs privacy-aware GWAS on distributed datasets while preserving the accuracy...

47. Spoki: A Reactive and Scalable Network Telescope (66)

Poster

Internet-wide scans are cheaply and quickly performed in IPv4. They are not only used to analyze the Internet ecosystem but abused to find vulnerable systems. We developed Spoki, a reactive-network telescope built on top of native actors in C++. It accepts TCP connections and collects payloads to look beyond the source addresses and get deeper insight into scanners.

Spoki is deployed at...

6. Studying protein 3D structures using network science (112)

Poster

Understanding protein 3-dimensional (3D) structures is important because functions of proteins depend on them. We proposed (static or dynamic) network approaches to model protein 3D structures as protein structure networks (PSNs). Static PSNs model the whole 3D structure of a protein as a single-layer PSN. Because the folding of a protein entails some protein parts folding before others, we...

30. Supporting Sustainable Research: The Services of the Center for Sustainable Research Data Management (15)

Poster

The Center for Sustainable Research Data Management (RDM Center) is a central institution of the Universität Hamburg that bundles infrastructures, competencies, and tasks to offer sustainable services to researchers.
The poster will provide an overview of what these services are and how they can be helpful in the everyday tasks researchers are faced with.
The services are, among others, the...

59. Sustainable data management @ Tamilex (158)

Poster

The Tamilex project, funded by a long-term grant from the Akademie der Wissenschaften in Hamburg, aims to create the first historical dictionary of Classical Tamil. It comprises an online corpus of annotated texts, linked to manuscript images, which provides the material evidence for lexical entries. Each of these layers requires careful consideration in order to ensure that the data is...

11. The AIMe registry for artificial intelligence in biomedical research (114)

Poster

Artificial intelligence's vast potential in biomedicine demands rigorous research with comprehensive method and data details. Despite journal peer review and checklists, variations in validation, metrics, data/code accessibility persist in publications, hampering comparability and reproducibility. The AIMe registry, a community-driven platform, enhances biomedical AI model accessibility,...

107. The CDCS - A Collaborative Research Center for Data Science in Hamburg (97)

Poster

The Center for Data and Computing in Natural Sciences (CDCS) is a joint facility of the University of Hamburg, the German Electron Synchrotron DESY and the Hamburg University of Technology. An increasing amount of scientific research projects rely on modern information technology. To accomodate for this development, the CDCS incorporates four Cross-Disciplinary Labs (CDLs) which are supported...

149. The Digital Causality Lab - New Ways to Teach Causality (35)

Poster

The Digital Causality Lab (DCL) is an innovative teaching project at the University of Hamburg which focuses on the topics Causal Inference and Data Literacy. It is one of the teaching projects that have been funded in the DDLitLab Project (Stiftung Innovation in der Hochschullehre). In the Digital Causality Lab, we use interactive apps to complement...

106. The DW-DGS: A corpus-based dictionary of German Sign Language is being compiled (118)

Poster

The Digitales Wörterbuch DGS (DW-DGS) is the first corpus-based dictionary of German Sign Language (DGS). As such, it supplies dictionary users with new types of information based on the analysis of real natural language data. This, for example, includes information on collocations and examples taken directly from the corpus. We show how the various target groups of the dictionary may benefit...

26. The Multilingual Sign Languages Wordnet (126)

Sam Bigeard (IDGS, Universität Hamburg)

Poster

Wordnets are a popular type of lexical resource, used in a variety of computational and linguistic applications. They are appreciated for their sense-based representation of lexical items and their typed relationship network. Wordnets for many different languages exist, but while research has been published toward creating Wordnets for sign languages, our work is the first resulting in a...

162. The Prompt-a-thon: Designing a Format for Value Co-Creation with Generative AI for Research and Practice (34)

Poster

Generative Artificial Intelligence (GenAI) has immense potential for innovation and problem solving, but there remains a significant knowledge gap in effectively harnessing its capabilities for value creation. To bridge this gap, a Design Science Research project was undertaken to develop a practical tool for researchers and practitioners to explore first insights into the usage of...

128. The Role of Digitalization for Firm Financial and Ecological Performance: A Socio-technical Systems Perspective (21)

Poster

Through the lens of Socio-Technical Systems theory, this study scrutinizes the impact of digitalization on both the financial and the ecological aspect of firm performance. We propose that digitalization, i.e., a significant modification in the technical subsystem of a firm, can amplify a firm’s financial and ecological performance only if it is accompanied with an equivalent shift in its...

24. The Sign Language Dataset Compendium: Creating an Overview of Digital Sign Language Resources (127)

Marc Schulder (IDGS, Universität Hamburg)

Poster

Identifying suitable datasets is a common challenge for data scientists working in domains with scarce data. For research on sign languages, this usually involves extensive literature review or word-of-mouth. Information on individual datasets may be distributed across different publications, data repositories and (potentially defunct) project websites. We introduce the Sign Language Dataset...

97. The VoicesCarry App: Co-Designing Worker Engagement Technology with Supply Chain Workers (50)

Poster

The Problem: businesses are digitally collecting information on workers’ wellbeing and labor condition in their supply chains, but little is known about the challenges that emerge from digitally-assisted auditing technology. Our poster describes the development of the VoicesCarry app, a web-based survey application that focuses on collecting self-reported information on supply chain workers’...

86. Towards the Automatic Generation of Models for Prediction, Monitoring, and Testing of Cyber-Physical Systems (62)

Poster

Modeling Cyber-Physical Systems (CPS) is a challenging task as they are composed of heterogeneous components that interact with each other and with the physical environment. This inherent
complexity requires a high level of knowledge about the system and its environment when modeling CPS. We develop a framework for the automatic generation of models for CPS, which integrate
into tools for...

135. Transformation of ERIS Information System to Advanced Search (149)

Poster

Our project focuses on developing the advanced search for ERIS information system which has been operated since 2012 by the Ancient History department at the University of Hamburg.

Since the current ERIS information system shows the database in text format, we aim to support historical researchers in finding hidden information in ancient biographical data by visualizing the data in charts....

50. Understanding and Improving Large Language Models (143)

Poster

The Machine Learning Group at UHH researches the core mechanisms behind large language models (LLMs). With the rise of ChatGPT, LLMs moved to the focus of public interest. Many research projects analyse the application of LLMs, while comparably little work sheds light on the algorithmic and mathemathical foundations behind training, reasoning and abstraction capabilities. We will advance LLM...

61. Using computational methods to study niche audiences (56)

Poster

This study employs computational methods, utilizing YouTube user data donations and survey data, to investigate the effects of fringe bubbles on social media. These niche communities, often situated at the margins of the public sphere, are shaped by algorithmically curated biased content, potentially distorting users’ perceptions of public discourse and amplifying non-mainstream voices....

37. Using PEMT for pharmaceutical patent landscaping (1)

Yojana Gadiya (Fraunhofer ITMP ScreeningPort)

Poster

Patents serve as critical catalysts in drug discovery, assuring legal safeguards to innovations while stimulating investments. By discerning patterns within patent data, researchers glean invaluable insights into pharmaceutical industry trends and priorities. Our study, powered by the Patent Enrichment Tool (PEMT), delves into patent literature concerning rare diseases (RD) and Alzheimer's...

19. Visual-Pattern Detection in Historical Documents (16)

Poster

Automatic pattern detection has become increasingly important for scholars in the humanities as the number of manuscripts that have been digitised has grown. Nevertheless, this task can be challenging for state-of-the-art computer vision approaches due to the lack of annotations and the small size of many patterns, which can be smaller than 0.1% of the image size. Therefore, we developed a...

110. Visualization and Sign Representation in the Digital Dictionary DGS (DW-DGS) (116)

Poster

Creating a dictionary of a visual language for diverse target groups presents unique challenges and calls for new technological solutions. We show ways of visualizing information, representing German Sign Language, and navigating the use of different languages in the Digitales Wörterbuch DGS (DW-DGS). As signs can best be represented in video form, we created a new format called a micon for...

43. When Recommender Systems Select Political News (73)

Poster

News recommender systems (NRS) determine news exposure for digital publics. However, the effects of NRSs on users remain understudied. This study examines how diversity of political news within NRSs can influence political news use, knowledge, and attitudes. We implemented an NRS that gathers news from a variety of news outlets, identifies political party representations in the articles, and...

82. World Data Center for Climate (WDCC) - Improvement of FAIRness of an established repository (100)

Poster

The World Data Center for Climate (WDCC), hosted by DKRZ, provides access to and offers long-term archiving for datasets relevant for Earth System research in a highly standardized manner. Its services are aimed at researchers who produce data and those who re-use published data for new research.

To meet user’s needs it is essential to ensure high quality of data, i.e. guaranteeing that...