Fakultät Informatik
Refine
Year of publication
Document Type
- conference proceeding (article) (173)
- Article (33)
- Part of a Book (20)
- Report (10)
- Bachelor-/Diplom Thesis (3)
- Book (3)
- Other (1)
- Working Paper (1)
Reviewed
- Begutachtet/Reviewed (204)
Keywords
- pathological speech (4)
- KI (3)
- Künstliche Intelligenz (3)
- neuropsychological tests (3)
- Artificial Intelligence (2)
- Automotive Security (2)
- ChatBot (2)
- Computerunterstützte Kommunikation (2)
- Graphdatenbank (2)
- Graphdatenmodell (2)
Institute
- Fakultät Informatik (244)
- Zentrum für Künstliche Intelligenz (KIZ) (55)
- Institut für e-Beratung (12)
- Fakultät Sozialwissenschaften (11)
- Fakultät Maschinenbau und Versorgungstechnik (2)
- Fakultät Betriebswirtschaft (1)
- Fakultät für Elektrotechnik Feinwerktechnik Informationstechnik (1)
- Institut für Angewandte Informatik (1)
- Institut für Chemie, Material- und Produktentwicklung (1)
- Lehr- und Kompetenzentwicklung (1)
Das Working Paper, welches im Rahmen des Projekts „CADS – Community Advisors – Digital Streetwork für EU-Beschäftigte in Deutschland“ erstellt wurde, untersucht, wie die Zielgruppe der EU-Zugewanderten durch digitale Beratungsangebote und E-Learning-Formate mithilfe von künstlicher Intelligenz (KI) effektiver begleitet, unterstützt und beraten werden kann. Dabei liegt der Schwerpunkt auf die Umsetzung der Lernformate auf sozialen Medienplattformen wie Facebook, Instagram und WhatsApp. Ziel ist es, praxisnahe Anwendungsszenarien und technische Lösungen für den Einsatz von KI-basierten Lerninstrumenten zu entwickeln, um die Integration von EU-Zugewanderten in den Arbeitsmarkt zu erleichtern. Ein besonderes Augenmerk liegt dabei auf der Einführung eines KI-gestützten Chatbots auf Facebook, welcher eine bedarfsorientierte und niedrigschwellige Unterstützung anbietet. Durch die Entwicklung solcher innovativen Ansätze sollen allgemeine Erkenntnisse zur Entwicklung von KI-basierten Lernformen gewonnen werden, die langfristig zur Verbesserung der Integration von EU-Zugewanderten in Deutschland beitragen können.
University students increasingly misuse generative AI (GenAI) to cheat on their academic assignments. Instructors need to develop pedagogic strategies which encourage active, authentic learning rather than just the superficial completion of tasks. A constructivist pedagogic concept which utilizes prompt engineering patterns to purposely integrate GenAI into an undergraduate course on databases is described. Students learn to develop structured queries to GenAI, based on prompt engineering patterns. This approach teaches inquiry-based exploration of complex topics in database design and development through iterative interaction with GenAI, which thus serves as an individually personalized tutor. Students learn to use GenAI as a cognitive tool, rather than just copying the answers provided. This approach makes use of scaffolding to empower students to master complex database principles.
Autonomous vehicles rely on high-quality sensor data, such as from camera and Light Detection and Ranging (LiDAR) for their driving decisions. Object detection and motion planning are core tasks in autonomous driving, both requiring robustness against diverse sensor perturbations. While robustness evaluation for object detection is well studied, far fewer approaches exist for end-to-end (E2E) motion planning agents. This paper outlines key differences between object detection and E2E motion planning and discusses resulting challenges for robustness evaluation. We present an empirical study comparing the robustness of the object detection system LoGoNet and the E2E agent TransFuser++, both using camera and LiDAR inputs, under LiDAR-only perturbations. Based on our findings, we propose requirements and a parameterization scheme for perturbations to guide the design of future robustness evaluation frameworks for autonomous agents.
This report examines the Examination Timetabling Problem (ETP) at the Department of Computer Science (IN) at Technische Hochschule Nürnberg (THN), focusing on the specific constraints and challenges of the institution. The report surveys various algorithms and approaches applied to solve the ETP. The Step Counting Hill Climbing (SCHC) algorithm, originally proposed
by Bykov and Petrovic, is highlighted as a particularly effective solution. The report details the implementation and enhancements of the SCHC algorithm at IN, which has been successfully applied to the department's timetabling needs. Experimental results demonstrate that the SCHC algorithm, with its simplicity and efficiency, produces high-quality timetables within minutes, comparable to manually created schedules. The report also discusses the limitations of integer linear programming for real-world ETPs due to its exponential complexity. The findings suggest that the SCHC algorithm is well-suited for the ETP at IN, with potential for further optimisation to improve
speed and solution quality.
Rising global demand for mental health support creates significant service delivery challenges, with asynchronous email counselling serving as a crucial low-threshold channel for accessing care. This paper presents CAIA, a co-designed AIbased tool suite that demonstrates responsible AI integration into counselling practice through seven LLM-driven functions enhanced by retrieval-augmented generation. A field evaluation involved 34 professional counsellors conducting authentic sessions with trained student counsellees (36 threads, 321 messages, 1,257 AI outputs). User behaviour analysis confirms substantial adoption, revealing that professional autonomy and information accuracy are decisive for sustained acceptance, with counsellors particularly valuing interpretive functionalities that provide new perspectives and stimulate professional reflection.
The expanding field of psychosocial online email counselling has created a demand for advanced tools to support counsellors in managing diverse client interactions. This paper introduces CAIA (Counsellor Artificial Intelligence Assistant), a comprehensive AI-driven system currently under active development, designed to condense and present information to counsellors in a structured manner. CAIA integrates several AI-based features, including Automatic Subject Generation, Case Summarisation, Timeline Management, Social Network Graph creation, and Case Classification. The features are designed to assist the counsellors, allowing them to engage more deeply with each case while ensuring that the client's needs remain paramount. Each feature presented in the paper addresses a specific problem in psychosocial online counselling, proposes a solution, discusses a technical implementation approach, and highlights the challenges encountered. Additionally, the paper discusses ethical considerations associated with an AI integration.
Large Language Models (LLMs) have the potential to enhance e-mental health and psychosocial e-mail counselling by automating tasks such as generating concise and relevant subject lines for client communications. However, concerns regarding accuracy, reliability, data privacy and resource efficiency persist. This study investigates the performance of several LLMs in generating subject lines for e-mail threads, yielding a total of 253 generated subjects. Each subject line was assessed by six raters, including five counselling professionals and one AI system, using a three-category quality scale (Good, Fair, Poor). The results show that LLMs can generally produce concise subject lines considered helpful by experts. While GPT-4o and GPT-3.5 Turbo outperformed other models, their use is restricted in mental health settings due to data protection concerns, making the evaluation of open-source models crucial. Among open-source models, SauerkrautLM LLama 3 70b (4-bit) and SauerkrautLM Mixtral 8x7b (both 8-bit and 4-bit versions) delivered promising results with potential for further development. In contrast, models with lower parameter counts produced predominantly poor outputs.
Deploying large language models (LLMs) in real-world applications requires robust safety guard models to detect and block harmful user prompts. While large safety guard models achieve strong performance, their computational
cost is substantial. To mitigate this, smaller distilled models are used, but they often underperform on “hard” examples where the larger model provides accurate predictions. We observe that many inputs can be reliablyhandled by the smaller model, while only a small fraction require the larger model’s capacity. Motivated by this, we propose SafeRoute, a binary router that distinguishes hard examples from easy ones. Our method selectively applies the larger safety guard model to the data that the router considers hard, improving efficiency while maintaining accuracy compared to solely using the larger safety guard model. Experimental results on multiple benchmark datasets demonstrate that our adaptive model selection significantly enhances the trade-off between computational cost and safety performance, outperforming relevant baselines.
Current work on speech-based dementia assessment focuses on either feature extraction to predict assessment scales, or on the automation of existing test procedures. Most research uses public data unquestioningly and rarely performs a detailed error analysis, focusing primarily on numerical performance. We perform an in-depth analysis of an automated standardized dementia assessment, the Syndrom-Kurz-Test. We find that while there is a high overall correlation with human annotators, due to certain artifacts, we observe high correlations for the severely impaired individuals, which is less true for the healthy or mildly impaired ones. Speech production decreases with cognitive decline, leading to overoptimistic correlations when test scoring
relies on word naming. Depending on the test design, fallback handling introduces further biases that favor certain groups. These pitfalls remain independent of group distributions in datasets and require differentiated analysis of target groups.