Förderung für Forschung mit Azure-Services

Organisational unit: Faculty of Computer Science, Research Group Security and Privacy

Abstract:

We will investigate how optimization techniques used for deploying AI models on edge devices affect their security against various types of attacks. Many real-world systems apply quantization and related methods to reduce model size and accelerate inference, but the security implications of these transformations are still not well understood. In this project, we will systematically compare full-precision models with quantized variants to measure how optimization changes both predictive performance and robustness against prompt injections and other types of attacks.

Our evaluations will cover different model classes, including large language models (LLMs) and vision-language models (VLMs), and will include extensive attack simulations under realistic threat assumptions. We will study trade-offs between accuracy, computational efficiency, and resilience against attacks, and we will analyze whether certain quantization settings or model families show consistent security patterns. The results will provide practical guidance for deploying smaller and more energy-efficient AI systems while maintaining reliable security properties in real-world environments.

Organisational unit: Department of Government

Abstract: 

This project examines the spatial distribution and strategic use of election posters during the 2025 German federal election, focusing on the district of Friedrichshain-Kreuzberg in Berlin. While previous research has analyzed poster design and messaging, the geographic placement of campaign posters has received limited attention due to the challenges of collecting and processing spatially detailed data. To address this gap, the project builds on a manually collected dataset of more than 10,000 georeferenced photographs of election posters. 

As an initial step, a random subset of 500 photographs has been manually double-coded, providing validated annotations on poster presence, recurring poster motifs, and substantive content categories. This manually coded subset serves as both a validation benchmark and a training resource for the next phase of the project, which focuses on scaling the analysis to the full dataset using automated methods. Specifically, the automated pipeline proceeds in three steps. First, election posters are detected and extracted from raw photographs using pre-trained computer vision models. Second, the extracted poster images are assigned to recurring poster motifs, reflecting repeated campaign designs, using image-based similarity measures and classification models. Third, the substantive content of each poster motif is classified, including party affiliation, issue focus, and degree of personalization. This final classification step combines classical machine-learning approaches with AI-supported methods, including image-based analysis (e.g., face detection to identify candidate-centered posters), text extraction via OCR, and natural language processing enhanced by LLM-based coding to interpret and classify textual elements on the posters. 

To implement this hybrid pipeline, the project employs a combination of offline local data collection and cloud-based computation in Microsoft Azure. Pre-trained computer vision models, classical machine-learning classifiers, and AI models (including LLMs) are applied for poster and face detection, motif recognition, and content classification. The largescale processing of high-resolution images and multimodal model application exceeds available local computing resources and cannot be efficiently supported by existing institutional IT services, making Azure’s scalable AI and compute services essential. The project’s innovative contribution lies in transitioning from manual coding to a largely automated, reproducible, and scalable workflow for analyzing offline campaign strategies. The project team is currently preparing a follow-up research proposal that aims to extend this approach to substantially larger image datasets in future election campaigns.

Organisational unit: Faculty of Business, Economics and Statistics, Department of Business Decisions and Analytics

Abstract:

Computing delivery routes, scheduling production lines, and allocating humanitarian supplies are all challenging combinatorial optimization problems that organizations must routinely solve. Traditionally, experts design algorithms through a time consuming process of trial and error. Recently, our research group developed methods that automatically generate optimization approaches, called heuristics, using large language models (LLMs) given a problem specification and data. To date, this work has focused on routing problems, such as package delivery. This project extends our technique to broader classes of optimization problems. It will employ iterative heuristic discovery powered by extensive LLM queries, exploring different models and search algorithms to produce high quality heuristics that match or surpass state of the art human designed methods. Our LLM based discovery process democratizes access to advanced optimization, enabling organizations to tackle complex problems without specialized consultants. Moreover, the heuristics produced by our approach deliver higher quality solutions in less time than competing human crafted techniques.

Organisational unit: Faculty of Business, Economics and Statistics

Abstract: 

How political elites and news media frame intergroup relations — as cooperative or competitive — has far-reaching consequences for polarization, public attitudes, and policy preferences. While surveys capture how individuals perceive intergroup dynamics, no systematic, high-frequency measure exists for the narratives they are exposed to. We launch a research agenda to construct the first cross-national index of how intergroup relations are framed in public discourse.

Using large language models, we develop a classification framework that captures both the nature and intensity of intergroup framing in political and media texts across countries, policy domains, and media formats. The resulting index will enable researchers to study how the framing of intergroup relations in public discourse covaries with — and potentially shapes — political outcomes, policy shifts, and economic shocks.

Organisational unit: Faculty of Psychology, Department of Occupational, Economic, and Social Psychology

Abstract:

Background: Health information is often communicated under severe time pressure in busy clinical settings and in language that exceeds many patients’ health literacy levels. As a result, patients frequently leave medical encounters uncertain about critical decisions affecting their health — including decisions about vaccination. Generative AI-chatbots powered by Large Language Models (LLMs) offer a promising way to extend access to clear, tailored, and on-demand medical information beyond medical settings. However, despite their rapid adoption, there is little evidence that currently available chatbots can support informed decisionmaking while reliably adhering to established health-communication guidelines.

Research set-up: By embedding AI-assisted communication within a controlled decision environment, this project moves beyond speculative claims and provides rigorous causal evidence on whether—and under what conditions—generative AI can meaningfully improve vaccination decision-making. To this end, we employ an established vaccination decision-making paradigm, in which participants face a realistic decision scenario involving a fictitious infectious disease for which a vaccine is available, requiring them to weigh disease risk against potential vaccine side effects. Participants’ choices carry tangible consequences within the experiment, affecting their accumulated “fitness points,” which are converted into monetary payoffs and serve as a proxy for health outcomes. To inform their decisions, participants are given access to a chatbot constrained to a carefully curated, scenario-specific information base. Specifically, using Azure Open AI services combined with Azure Container Apps, a GPT-4o model will be trained on epidemiological information regarding the disease, the vaccine, and the broader decision context.

Research program: In Study 1, we evaluate whether a generative AI-chatbot prompted to use simple, plain language can improve users’ understanding and decision-making in a vaccination context. Study 2 moves beyond language alone by testing how both user framing and chatbot behavior shape outcomes: participants are either instructed to interact with the chatbot as if it were a human health professional or not, and the chatbot is either instructed to follow established medical communication guidelines or not. Across both studies, we assess outcomes central to effective health communication, including trust in the chatbot, knowledge about the disease and vaccination, decisional conflict, and perceived usability. Crucially, we also directly evaluate the chatbot’s adherence to simple language and medical communication guidelines, respectively. By systematically isolating these design factors, this project generates much-needed empirical evidence on how generative AI can—and cannot—be configured to support vaccination decision-making.

Organisational unit: Department of Government

Abstract:

This project examines Austria’s unusually large and hard to oversee body of constitutional legislation, exploring its causes and consequences. By integrating comprehensive data from the Austrian Legal Information System, parliamentary records, and Constitutional Court materials, the project aims to explain how political actors choose between ordinary and constitutional legislation and how these choices shape long-term policy constraints. While related questions have often been addressed descriptively, largely by scholars of law, this project advances the field by applying quantitative research designs and by analyzing Austria’s constitutional corpus from a political science perspective. In earlier project phases, we assembled several datasets covering all federal laws, the universe of constitutional provisions, and all cases of constitutional review in Austria since 1945. Because these data originate from a variety of sources and span multiple decades, much of the material is unstructured and lacks consistent metadata. Recently, we successfully applied a Large Language Model (LLM) to classify all constitutional-rank articles into policy areas, enabling us to track changes in constitutional regulatory density in different policy areas over time. Currently, we are starting to analyze all legislative initiatives in Austria since 1945 – nearly 20,000 initiatives available as machine-readable files but, for most of the period, without meaningful metadata. The objective is to analyze the factors that determine whether an initiative is adopted with constitutional rank, enacted as ordinary legislation, or fails in the legislative process. Additional funding for Microsoft Azure, which we would use on OpenAI in Azure, would enable us to use it to categorize these initiatives into different policy areas and enhance them with additional variables necessary for our analyses.

Organisational unit: Faculty of Informatics, Research Group Security and Privacy

Abstract: 

Website phishing remains one of the most widespread and damaging forms of online fraud, enabling attackers to steal user credentials, financial data, and other sensitive information by imitating legitimate online services. As attackers increasingly adopt realistic visual designs (e.g., logos, layouts, branding elements) alongside adaptive content and dynamic behaviors, traditional rule- or URL-based filters are becoming less effective. Conventional machine-learning approaches for phishing detection rely on extensive feature engineering and require frequent retraining.

This project aims to assess the potential of generative AI, particularly multimodal models, to reason jointly over textual, structural, and visual inputs to detect malicious websites. Such models reduce the need for handcrafted features and offer more adaptive and generalizable detection mechanisms. We will systematically benchmark a diverse set of open-source (e.g., Llama, DeepSeek, Qwen, Gemma) and proprietary (e.g., GPT, Grok) (vision) language models (LLMs/VLMs) for phishing website detection and classification. The evaluation will cover multiple modalities such as HTML content, URLs, and website screenshots, and employ publicly available datasets comprising tens of thousands of samples.

While locally deployed (smaller) vision models provide advantages in control, adaptability, and cost predictability, large cloud-scale models typically offer superior performance and scalability. To quantify these trade offs, we will conduct extensive experiments under several configurations:

1. Mixture-of-experts setups, combining complementary LLM/VLM models,
2. Hybrid integration with classical ML approaches, using OCR and image-based
   feature extraction,
3. Combining LLMs/VLMs with classical ML, where LLM outputs serve as features for
   downstream classifiers, and
4. Knowledge-augmented reasoning, incorporating external sources such as RAG and
   GraphRAG.

Selected models will be fine-tuned on phishing-specific data, with all experiments incorporating trace logging and transparent documentation to ensure reproducibility, interpretability, and fair benchmarking. The project will systematically assess performance, cost, and scalability trade-offs, laying the foundation for advanced AI-assisted cybersecurity research. All results, including detailed setup instructions, will be published open access.

Organisational unit: Faculty of Physics, Research Group Nanomagnetism and Magnonics

Abstract: 

Spin-wave dispersion analysis is the fundamental first step in characterizing magnetic eigenmodes in a given material and geometry. It covers a broad range of research fields, including quantum magnonics and microwave spintronics. Importantly, dispersion calculations also form the basis for developing complex magnonic data-processing units, including inverse-designed devices, an approach increasingly relevant for AI inspired and neuromorphic computing [1]. Despite its importance, such simulations typically require specialized expertise and locally installed software, limiting accessibility and reproducibility. 

This project aims to develop and operate an open-access, cloud-based spin-wave dispersion calculator that allows users to configure, run, and analyze simulations through a web interface. The software will be actively used by two research units at the University of Vienna: the Nanomagnetism and Magnonics group (Univ.-Prof. Dr. Andrii Chumak) and the Physics of Functional Materials group (Univ.-Prof. Dr. Dieter Suess). It will be also openly accessible to external collaborators and researchers worldwide, in line with modern principles of open academic knowledge exchange. The platform is implemented using Microsoft Azure services. An Azure Static Web App provides the frontend for defining material parameters, geometries, and excitation conditions [2, 3]. Simulations are executed in an Azure Container App using the open-source TetraX micromagnetic solver [4] to compute spin-wave dispersion relations [5], while simulation data and logs are stored via Azure File Share to ensure reproducibility. 

During the funding period, the platform will be extended beyond basic dispersion calculations toward more complex simulations of magnetization dynamics and collective excitation phenomena. These developments will require increased computational resources and will utilize Azure Virtual Machines for scalable CPU and GPU computing. In parallel, container orchestration will be migrated to Azure Kubernetes Service (AKS) to improve resource management and scalability for multiple concurrent users. By combining cloud-based infrastructure with a well-established opensource simulation engine, MaDiVie will provide a reliable, extensible, and reproducible research tool for the international magnonics community.
 

References

[1] Chumak, A. V., et al., Advances in magnetics roadmap on spinwave computing, IEEE Transactions on Magnetics 58, 1–72 (2022). 

[2] Frontend website: https://www.madivie.at  

[3] Frontend GitHub repository: https://github.com/GIGAluckman/WebDispersionCalculator 

[4] Körber, L., et al., Finite-element dynamic-matrix approach for spin-wave dispersions in magnonic waveguides with arbitrary cross section, AIP Advances 11, 095006 (2021). 

[5] Backend GitHub repository: https://github.com/GIGAluckman/WebDispersionCalculator-backend 

Organisational unit: Vienna University Library and Archive Services, Library - Research and Publication Services

Abstract:

The iConTxt project brings a transformative innovation to the Digital Library and Bibliography for Literature in Translation and Adaptation (DLBT) by integrating advanced artificial intelligence to unlock the full potential of its unique multilingual collections. As one of the most comprehensive and authoritative resources on literature in translation, the DLBT holds vast scholarly value, yet much of this material remains underused due to language and accessibility barriers. iConTxt directly addresses this challenge. Developed collaboratively by eight European universities in partnership with the Vienna University Library, the project leverages Large Language Models (LLMs) to automate metadata curation, generate English-language versions and summaries of reception materials, and produce new contextual knowledge derived from existing database materials. In this way, iConTxt transforms isolated records into interconnected, research-ready information.

By embedding AI capabilities directly into the YARM software that powers the DLBT, iConTxt transforms the library into a dynamic, self-enriching research infrastructure. It enables user-friendly access for researchers, students, translators, literary agents, cultural institutions, and policymakers, fostering new opportunities for cross-cultural analysis, educational innovation, and literary exchange. Its open-source design ensures that the resulting tools and methodologies remain freely available to registered users. It enhances global accessibility to cultural heritage, makes the trajectories of translated literature more visible, and contributes to a broader understanding of literature as a living, transnational ecosystem.

Organisational unit: Faculty of Business, Economics and Statistics - Vienna Center for Experimental Economics

Abstract: 

In economics experiments, participants who are confused about a task’s rules often act as though they held a preference. That confusion is then mistaken for the preference itself, which distorts the conclusions and policy advice drawn from it. This project asks a simple question: can a chatbot reduce such confusion by clarifying the rules for participants, without ever biasing their choices? Using Microsoft Azure AI Foundry, we built such an assistant. To test how well it stays within its limits, five kinds of adversarial participant engage it over whole conversations and try to make it misbehave, while an automated judge scores every reply for three failures: giving advice, saying something factually wrong, and reducing the task’s complexity rather than just explaining its rules. We ran this over six assistant designs, three models (nano, mini, and full), and two payoff scenarios, about 3,400 conversations and over 50,000 scored replies in all. One result stands out. The model matters far more than the architecture: paired with the simplest design, a capable model reduces all three failures to almost nothing, with advice leakage well under one percent. The more elaborate multi-agent designs add cost and delay without improving on this. We recommend the simplest design, running on a capable but inexpensive model. For more detail and this work’s continuing development into a general testing tool, see our project page.

Organisational units:

• Vienna University Library and Archive Services – Department for Bibliometrics and Publication Strategies,
• Vienna University Library and Archive Services – Library-Communications,
• Zentraler Informatikdienst – IT-Support for Research

Abstract:

Project SciTextMatch explores how AI-supported text analysis can help identify promising research collaboration opportunities between universities. The project focuses on comparing research profiles and publication information from selected principal investigator (PI) candidates in order to highlight potential matches across institutions. 

During the reporting period, a Python-based prototype workflow was developed and tested for the strategic identification of collaboration opportunities, particularly in relation to international partnerships and upcoming Horizon Europe calls. The approach combines publication-based similarity analysis with AI-supported validation and review steps. 

The prototype has been successfully implemented and applied in a real-world strategic-partnership context, with a focus on identifying potential collaboration opportunities between the University of Vienna and the University of Toronto. 

Organisational unit: Centre for Teacher Education – Department for Teacher Education

Abstract:

The project developed a contextualized chatbot for the CE-Lab to provide equipment information, troubleshooting, and booking support directly within the lab’s booking platform. Objectives were met through structured use-case analysis, rapid prototyping of tailored AIs using Dify and Azure OpenAI, domain-specific contextualization via curated CE-Lab documents, and iterative scenario testing with pilot feedback to refine prompts and system settings. Key results include a stable, scalable chatbot embedded in test mode with single-resource booking, defined workflows for multi-resource booking and automatic detection of ancillary resources, and validation of model suitability (GPT and Mistral evaluated via Open WebUI). A major challenge – automated ingestion of relevant content (Firecrawl) – was addressed by establishing a manual curation and update process, improving answer quality and reliability. The primary goal was achieved; the secondary goal (didactic recommendations) was partially advanced and planned for follow-up work. The outcomes reduce staff workload, enhance student support, and lay a foundation for extensible educational and research applications.

Organisational unit: Faculty of Informatics – Research Group Security and Privacy

Abstract:

High-stakes sociopolitical events such as elections, social movements, and terrorist attacks are often accompanied by the expressions of perceived injustice – commonly framed as grievances. In this project, we investigate how grievances manifest linguistically and how they contribute to violent intent and shape community formation in online social networks. We analyze a dataset comprising 76.7 million tweets related to high-stakes events and manifestos written by confirmed violent offenders. By relying on psycholinguistic analysis, graph-based embeddings, attributed community detection, and large language models, we examined grievances across multiple textual genres and temporal social networks. We show that violent authors’ texts form distinct and cohesive communities characterized by their heightened expressions of aggression, moral absolutism, and expression of grievances. Ablation analyses revealed that grievances acted as a cohesive bond rather than serving as a discriminative feature.

By extending this analysis to social media interactions, we showed that including grievances as contextual markers into multiplex temporal networks increases community cohesion during high-stakes events. Finally, we introduced and validated a fine-grained taxonomy of grievances based on interdisciplinary theory and survey responses.

We offer methodological and theoretical insights for the study of high-risk discourse – including radicalization and polarization.

Organisational unit: Faculty of Psychology – Department of Cognition, Emotion, and Methods in Psychology

Abstract:

Building and Evaluating Historical Large Language Models for Understanding Psychological Dimensions in Fiction investigates how modern AI can be adapted to analyze psychological themes in historical European fiction (16^th–19^th centuries). We developed two complementary approaches: 

1. prompt-based annotation using Azure OpenAI (GPT-4) to classify text segments into six broad virtue clusters (knowledge, courage, humanity, justice, temperance, transcendence) and
2. groundwork for fine-tuning open-source large language models on historical corpora.

During the project, we built a reusable pipeline to parse early modern theatre texts, extract speech-level units with metadata, and annotate them with structured, historically informed prompts. We also constructed and validated historical semantic embedding spaces (fastText + PPMI) for English and French to support future model adaptation. As an empirical benchmark, we compared LLM-based play-level virtue profiles with independent dictionary-based measures (LIWC-style bag-of-words). The two approaches showed strong convergence for several domains (especially knowledge, humanity, and justice) and weaker convergence for others (temperance and transcendence), highlighting where contextual AI methods add value beyond word-count approaches.

The project establishes validated tools and benchmarks for future large-scale annotation and targeted fine-tuning of Historical Large Language Models using Azure GPU resources.

Organisational unit: Faculty of Business, Economics and Statistics – Department of Accounting, Innovation and Strategy

Abstract:

The project aimed to develop an intelligent AI student assistant chatbot for university seminars that can answer both organizational and content-related questions. The solution was designed around a syllabus-centric knowledge base and natural language interaction. In an initial Azure-based prototype, Azure AI Search indexed an updatable syllabus stored in Azure Blob Storage, while Azure OpenAI (GPT-4 deployment) generated answers to student queries. This demonstrated that syllabus-driven assistance is technically feasible and can relieve instructors of routine Q&A. However, practical challenges with Azure’s complexity, regional service limitations, and deployment tooling led to a strategic decision at mid-term to stop extending the Azure implementation. For further experimentation and teaching use, the assistant concept was subsequently pursued with alternative, lower-friction platforms (e.g. ChatGPT Agents, OpenAI platform).

Organisational unit: Faculty of Historical and Cultural Studies – Department of European Ethnology

Abstract:

This pilot project explored how computational text analysis can be applied from a linguistic-anthropological perspective to the study of AI governance and regulatory language. Focusing on EU and US AI policy documents, the project developed and tested an exploratory workflow combining document preprocessing, sentence-level extraction, semantic embeddings, and unsupervised clustering. Due to policy and authentication constraints affecting cloud-based model deployment, the analysis relied on locally executed open-source embedding models within an Azure Machine Learning environment. The results show that EU and US regulatory texts enact distinct “ontologies” of AI, not primarily through abstract definitions, but through recurring patterns of classification, administrative procedure, risk management, and human impact framing. The project thus provides a proof of concept for integrating computational methods with interpretive linguistic-anthropological analysis and yields methodological insights that directly inform the design of a larger follow-up research project.