Research into artificial intelligence is not a very young field; its precursors can be traced back as far as the 16th century. Today’s technical development, however, is virtually leaping forward, with new intelligent chat systems and social robots playing no small part in this. This is revolutionizing a wide range of scientific and social fields. The very large publication numbers in this field illustrate this as well. In order to keep track of the discourse in the field, the representatives of the field, the publications as well as the topics and their future development, it is indispensable for academics and scientists to prepare them in a bibliometric analysis. Only in this way it is possible to uncover thematic gaps as well as further points of contact and to drive research forward in a targeted and stringent manner. It is precisely this sorting and processing of the research discourse, the topics, and the authors, which is necessary for further research, that is carried out in this paper. For this purpose, using bibliometric analysis tools, an overview of the past, present, and future of the research field is created, and the general relevant topics are uncovered. The analysis includes as performance analysis a) the total number of publications and b) the total number of citations, and for science mapping c) a co-citation analysis (past), d) a bibliographic coupling (present) and e) a co-word analysis (future). The data needed for the analysis are identified and extracted from the SCOPUS or Web of Science (ISI) databases

“China expertise” has become the keyword in European foreign politics. A look at the latest government publications shows that a clear definition is yet to be expected. It can be summarized from reports about educational programs that there are differing trends among European countries as regards their Chinese language and other China-related skills as well as their infrastructures to gain these. A linguistic assessment shows that for learning this distant language, an early onset of acquisition is recommended. Given the almost pan-European desire for China expertise a clear definition is demanded so efficient international cooperation to build up this expertise over-regionally is possible.

Ongoing assessments in a course are crucial for tracking student performance and progress. However, generating and evaluating tests for each lesson and student can be time-consuming. Existing models for generating and evaluating question-answer pairs have had limited success. In recent years, large language models (LLMs) have become available as a service, offering more intelligent answering and evaluation capabilities. This research aims to leverage LLMs for generating questions, model answers, and evaluations while providing valuable feedback to students and decentralizing the dependency on faculty. Our approach is based on the development and interaction of advanced AI-powered generative agents, built on large language models like ChatGPT, GPT-4, and Vicuna, and designed to emulate human activities such as information abstraction, context refinement, and query rating. These agents interact autonomously in a network, employing techniques like Zero-Shot and Few-Shot Prompting to generate responses and adapt to various roles and contexts. The setup includes three key agents for question generation, refinement, and quality assurance, which leverage text vectorization, document selection and filtering, and cutting-edge language models to generate, refine, and evaluate questions and answers based on specific learning objectives. In conclusion, this paper demonstrates the versatility of LLMs for various learning tasks, including question generation, model answer generation, and evaluation, all while providing personalized feedback to students. By identifying and addressing knowledge gaps, LLMs can support continuous assessment and help students improve their understanding before semester exams.

This study demonstrates the need for novel gas engine control systems for com-
bined heat and power plants, also known as cogeneration power plants, connected to natural
gas grids. Hydrogen addition to natural gas grids in a range of up to 5% by volume is already
permitted throughout Europe. This offers the possibility to reduce carbon dioxide emissions
of end consumers connected to public natural gas grids and contributes to climate protec-
tion. However, conventional engine controls are not designed for natural gas/hydrogen mixture
operation. We tested fuels with up to 30% hydrogen by volume using a commercial six-cylinder
spark ignition engine, designed for natural gas or biogas operation in power plants. With engine
settings according to usual cogeneration operation, nitrogen oxide emissions increased expo-
nentially with increasing hydrogen amounts. We demonstrate that the usual approach of using
the lower heating value of the fuel mixture to regulate the engine is unable to accommodate the
hydrogen induced changes. For this reason, we developed a mathematical model to determine
the nitrogen oxide emissions based on boost pressure and power output. The idea behind this
novel approach is to regulate the engine based on emissions, regardless of the fuel gas. In this
work the approach for this virtual sensor is described and its performance demonstrated.

The wave of digitization has begun. Organizations deal with huge amounts of data, such as logs, websites, and documents. A common way to make the information contained in these sources machine-accessible for automated processing is to first extract the information and then store it in a knowledge graph. A key task in this approach is to recognize entities. While common named entity recognition (NER) models work well for common entity types, they typically fail to recognize custom entities. Custom entity recognition requires data to be manually annotated and custom NER models to be trained. To efficiently extract the information, this paper proposes an innovative solution: Our Gazetteer approach uses a knowledge graph to create a coarse and fast NER component, reducing the need for manual annotation and saving human effort. Focusing on a university use case, our Gazetteer is integrated into a chatbot for entity recognition. In addition, data can be annotated using the Gazetteer and an NER model can be trained. Subsequently, the NER model can be used to recognize unseen custom entities, which are then added to the knowledge graph. This will improve the knowledge graph and make it self-extending.

Extracting triples of subjects, objects, and predicates from text to populate knowledge bases traditionally involves several intermediate steps such as co-reference resolution, named entity recognition, and relationship extraction. Treating triple extraction as translation task from source sentences to sets of triples, we present an end-to-end solution for information extraction that uses task prefixes to prompts a fine-tuned large language model to extract triples from text. Thus, the need for data labeling and training multiple models is reduced.

Individuals and groups are frequently targets of bullying, sexual harassment, and hate speech on online platforms. Such norm violations can have detrimental negative consequences, for instance by causing psychological harm and damaging social cohesion. Finding ways to reduce and prevent online norm violations is hence crucial. Online users may play an important role in this context. We argue that it can be considered morally courageous if users decide to take a stand against perceived violations of their own moral beliefs and endorsed norms, as it may imply substantial risks for themselves. With this chapter, we aim to advance our understanding of online moral courage as a relatively new phenomenon. First, we provide an examination of critical characteristics of online environments that may facilitate or hinder moral courage. Second, we discuss consequences of online moral courage by considering its effects on perpetrators, further online users, and the general tonality of the online discourse. Last, we integrate insights on the facilitators and obstacles of online moral courage and its consequences to provide practical recommendations for the design and management of online platforms and user education and training.

This paper presents the ongoing development of HAnS (Hochschul-Assistenz-System), an Intelligent Tutoring System (ITS) designed to support self-directed digital learning in higher education. Initiated by twelve collaborating German universities and research institutes, HAnS is developed 2021–2025 with the goal of utilizing artificial intelligence (AI) and Big Data in academic settings to enhance technology-based learning. The system employs AI for speech recognition and the indexing of existing learning resources, enabling users to search and compile these materials based on various parameters. Here, we provide an overview of the project, showcasing how iterative design and development processes contribute to innovative educational research in the evolving field of AI-based ITS in higher education. Notwithstanding the potential of HAnS, we also deliberate upon the challenges associated with ensuring a suitable dataset for training the AI, refining complex algorithms for personalization, and maintaining data privacy.

The following paper presents the evaluation of an artificially intelligent assistant system (DIAS) with a service-oriented chatbot as a central communication element. The conversational AI (Artificial Intelligence) is supposed to increase information transparency in higher education environments and thus support students, teachers, and administrative staff. The exploratory study had two objectives: first, we intended to find out about the usability and utility of the DIAS chatbot using the CUQ (Chatbot Usability Questionnaire) score and benchmark the results against other conversational agents. Secondly, we were interested in possible effects among the different variables of interest, which could contribute to further theory development of chatbots in education. The results show that the DIAS chatbot scored above average, and can support students in finding relevant information, particularly if they use the assistant frequently. Positive aspects included the intuitive use, a welcoming persona (expressed in design & language) and easy navigation. The negative feedback showed potential for improvement particularly in content quality and handling dialogue mistakes, which is a general shortcoming of conversational AI at this development stage. The results can be used as a guidance for future research and theory building. However, they must be considered carefully due to several study limitations.

In a current research project at the Ansbach University of Applied Science, an AI-based quiz function was created to serve as a voluntary student-oriented support offer to determine their learning progress in their respective courses by means of conducting self-assessment quizzes. The application takes lecture scripts as input and applies a question generation model to create questions that students can answer. In order to evaluate the given answers, another language model is involved to perform Natural Language Inference (NLI). Users can engage with the system via a graphical user interface currently provided via a web app. To assess preliminary feasibility and perception of the model prototype, a qualitative focus group discussion following a semi-structured interview guideline prepared by the research team according to similar studies in the education field (Sek et al. 2012) was conducted with five participants. A transcript of the discussion was prepared and analyzed using the qualitative content analysis method according to Kuckartz. Overall, the quiz function was well received by the participants of the focus group. However, the prototype still has potential when it comes to generating meaningful questions and transparently assigning categories to the given answers. Furthermore, the quiz parameters should be individually adjustable by users. In the following paper, the development of the service is illustrated by outlining the considerations for the application design and the training procedure of the language models. Afterwards, the design of the qualitative focus group is described including the presentation of the results.

Dieses essential gibt einen Überblick über verschiedene Go-to-Market-Strategien für zweiseitige digitale Plattform-Geschäftsmodelle. Zunächst liegt der Fokus auf den Strategien zur sequenziellen, simultanen und viralen Kundenakquise. Im zweiten Teil zeigt der Beitrag, mit welchen Faktoren man positive Interaktionen fördert, um eine kritische Masse an Nutzern zu gewinnen.