Large language models (LLMs) have garnered significant attention, but the definition of “large” lacks clarity. This paper focuses on medium-sized language models (MLMs), defined as having at least six billion parameters but less than 100 billion. The study evaluates MLMs regarding zero-shot generative question answering in German and English language, which requires models to provide elaborate answers without external document retrieval (RAG). The paper introduces an own test dataset and presents results from human evaluation. Results show that combining the best answers from different MLMs yielded an overall correct answer rate of 82.7% which is better than the 60.9% of ChatGPT. The best English MLM achieved 71.8% and has 33B parameters, which highlights the importance of using appropriate training data for fine-tuning rather than solely relying on the number of parameters. The best German model also surpasses ChatGPT for the equivalent dataset. More fine-grained feedback should be used to further improve the quality of answers. The open source community is quickly closing the gap to the best commercial models.

Class diagrams are at the core of object oriented modeling. They are the foundation of model-driven software engineering and backed up by a wide range of supporting tools. In most cases, source code may be generated from class diagrams which results in increasing productivity of developers. In this paper we present an approach that allows the automatic conversion of hand-drawn sketches of class diagrams into corresponding UML models and thus can help to speed up the development process significantly.

Large-scale attacks on IoT and edge computing devices pose a significant threat. As a prominent example, Mirai is an IoT botnet with 600,000 infected devices around the globe, capable of conducting effective and targeted DDoS attacks on (critical) infrastructure. Driven by the substantial impacts of attacks, manufacturers and system integrators propose Trusted Execution Environments (TEEs) that have gained significant importance recently. TEEs offer an execution environment to run small portions of code isolated from the rest of the system, even if the operating system is compromised. In this publication, we examine TEEs in the context of system monitoring and introduce the Trusted Monitor (TM), a novel anomaly detection system that runs within a TEE. The TM continuously profiles the system using hardware performance counters and utilizes an application-specific machine-learning model for anomaly detection. In our evaluation, we demonstrate that the TM accurately classifies 86% of 183 tested workloads, with an overhead of less than 2%. Notably, we show that a real-world kernel-level rootkit has observable effects on performance counters, allowing the TM to detect it. Major parts of the TM are implemented in the Rust programming language, eliminating common security-critical programming errors.

This book summarizes studies conducted by the Empirical Research and User Experience (ERUX) research group at the Institute for Information Systems (iisys) at Hof University of Applied Sciences over the period 2019 to 2022. The main topic focuses on “The Human-Machine-Interface in Advanced Driver Assistance Systems towards Autonomous Driving”, investigated using the example of automatic parking assistant in cars. A broad spectrum of methods was used to ensure comprehensive analyses and valid results. Starting situation for the studies has been the industry's realization that the demand for parking assistants had fallen short of expectations. Moreover, they concluded that a host of driving assistant system owners´ do not use them. This indicates an unsatisfactory acceptance of new technologies on the part of car drivers. The aim of the studies was therefore to determine the reasons for this restraint, as well as to derive measures to increase acceptance of this systems among end customers.

The first study focused on operators and the restraint of German drivers towards driver assistance systems. The results of this quantitative study displayed that although the modernity of these systems has been identified, the functional enhancements they offer are insufficient. In order to overcome this discrepancy, manufacturers are advised to emphasize the functional advantages more strongly when communication with the end customer. Study also suggested that direct experience can significantly increase customers' willingness to purchase such a system. Therefore, it is of pivotal importance to introduce customers to the new technologies in a targeted manner.

Using conjoint measurement, a second study illustrates driver assistance systems already representing a significant benefit when purchasing a car, even though the trust in fully autonomous vehicles is still restricted. Drivers favor established car manufacturer in terms of the development of autonomous vehicles. The third study utilizes eye-tracking to examine operating patterns of parking assistance systems. Moreover, it derives potential for improvement in terms of intuitive design and user experience. Fourthly, alternative scenarios for the design of parking assistants are evaluated in a concept test. This indicated the vital importance of simple and acquainted operating options.

Across all methods, the customer actually being able to experience driving assistance systems is a decisive matter for the acceptance of these new technologies. Therefore, manufacturers should emphasize their focus on getting customers in touch with these systems, as well as offering them targeted usage options. In addition, the integration of customer feedback into development efforts is essential to produce intuitive, user-friendly user interfaces and thus promote the diffusion process of new technologies.