Search

Background: The utilization of AI language models in education and academia is currently a subject of research, and applications in clinical settings are also being tested. Studies conducted by various research groups have demonstrated that language models can answer questions related to medical board examinations, and there are potential applications of these models in medical education as well. Research question: This study aims to investigate the extent to which current version language models prove effective for addressing medical inquiries, their potential utility in medical education, and the challenges that still exist in the functioning of AI language models. Method: The program ChatGPT, based on GPT 3.5, had to answer 1025 questions from the second part (M2) of the medical board examination. The study examined whether any errors and what types of errors occurred. Additionally, the language model was asked to generate essays on the learning objectives outlined in the standard curriculum for specialist training in anesthesiology and the supplementary qualification in emergency medicine. These essays were analyzed afterwards and checked for errors and anomalies. Results: The findings indicated that ChatGPT was able to correctly answer the questions with an accuracy rate exceeding 69%, even when the questions included references to visual aids. This represented an improvement in the accuracy of answering board examination questions compared to a study conducted in March; however, when it came to generating essays a high error rate was observed. Discussion: Considering the current pace of ongoing improvements in AI language models, widespread clinical implementation, especially in emergency departments as well as emergency and intensive care medicine with the assistance of medical trainees, is a plausible scenario. These models can provide insights to support medical professionals in their work, without relying solely on the language model. Although the use of these models in education holds promise, it currently requires a significant amount of supervision. Due to hallucinations caused by inadequate training environments for the language model, the generated texts might deviate from the current state of scientific knowledge. Direct deployment in patient care settings without permanent physician supervision does not yet appear to be achievable at present. [ABSTRACT FROM AUTHOR]

The book "German as a Foreign Language: An Introduction" by Dietmar Rösler has been published in a revised edition. The new edition takes into account changes in the areas of technology, immigration, and online learning that have affected the field of German as a foreign and second language. The book covers topics such as the use of digital media in the process of learning German, the gender debate, and the needs of specific learner groups. It also emphasizes the importance of research methodology and the selection of teaching content in the context of globalization and migration. Various aspects of the field of German as a foreign language are presented, with particular emphasis on clear language. It is emphasized that there is no universal teaching method and that decisions in the classroom should be tailored to the needs of the learners. Other relevant aspects of language learning and teaching are also addressed, such as teaching and learning methods, skills, error treatment, and the integration of literary texts. It is emphasized that learning a foreign language is closely related to engaging with a foreign culture. It is noted that the integration of digital learning is not yet sufficiently discussed in this introduction and that the integration of artificial intelligence and natural language processing brings opportunities and challenges for foreign language teaching. [Extracted from the article]

The German Society for Information (DGI) is celebrating its 75th anniversary with a symposium and an evening event in Frankfurt am Main. The program includes keynotes by Prof. Dr. Rainer Kuhlen and Matthias Razum, as well as lectures, workshops, and discussions on various topics such as information and disinformation, information literacy, and artificial intelligence. In addition, the results of two projects on digitization and terminology will be presented. The DGI general assembly will take place after the symposium. The DGI is opening its mailing list to interested parties and has established a specialist group to work on its history. Online seminars on various topics are also being offered. The University of Hamburg offers various training sessions and workshops on topics such as artificial intelligence, knowledge management, and project planning. The TransferAgency of the University of Hamburg deals with the transfer of knowledge and information to various sectors, including the economy, politics, and civil society. Johann Herlyn is a speaker for information and documentation at the TransferAgency and is responsible for the systematic recording of data and activities in the transfer process. The TransferAgency is working on developing new methods and procedures for documentation, such as the use of natural language processing. [Extracted from the article]

Semantic software technologies transform large quantities of structured and unstructured data into computer-readable and computer-usable knowledge networks. The creation and maintenance of knowledge networks are achieved through agile interaction between humans and computers. With customized editors, Natural Language Processing, semantic classification and self-learning processes, context and text understanding gradually and increasingly mature. Such knowledge networks deliver explainable and comprehensible results. They are of great value for research and companies in every industry and open the door to innovative applications both within companies and for customers: Examples include editorial systems and product information systems with content and formal support for new digital information services and products, 360° input for business decisions and search and aggregation of semantic connections in existing content. [ABSTRACT FROM AUTHOR]

The bachelor thesis presented in this article deals with the results of a shared task in the field of eHealth. The aim is to improve the classification accuracy of selected clinical coding systems by using ensemble methods. Decisive factors are the values of the evaluation measures Mean Average Precision and F1-score. [ABSTRACT FROM AUTHOR]

Die künstliche Intelligenz (KI) und die zugrunde liegenden Methoden des maschinellen Lernens und der neuronalen Netzwerke haben in den letzten Jahren dramatische Fortschritte gemacht und Leistungen in Domänen erreicht, die bis vor kurzem als spezifisch menschlich und für Computer nicht zugänglich galten. In diesem Überblick werden die diesen Fortschritten zugrunde liegenden methodischen Entwicklungen kurz dargestellt und in der Folge aktuelle und potenzielle Anwendungen auf die Psychiatrie in drei Bereichen diskutiert: Präzisionsmedizin und Biomarker, Verarbeitung natürlicher Sprache und KI-basierte psychotherapeutische Interventionen. Abschließend wird auf einige Risken dieser neuen Technologie hingewiesen. [ABSTRACT FROM AUTHOR]

With networked large data collections and suitable quantitative methods new research questions become possible. Within the DARIAH-DE cluster 5 new philologico- critical and historical comparing perspectives enter the modelling of large data collections. Consequently new, tailored analytical tools can be developed. This includes the chance to test in empirical ways theoretical presuppositions and the extension and advancement of the modes of the production of scientific knowledge. The article describes two analytical tools developed within cluster 5: the Dariah-DKPro-Wrapper, a software package for the automated linguistical analysis, and Cosmotool, a searchand analysis tool for the exploration of transboundary lives by means of structured and unstructured data collections. [ABSTRACT FROM AUTHOR]

The article discusses a workshop about digital data in the humanities held at the Bavarian Academy of Sciences and Humanities in Munich, Germany, from January 28-29, 2016. The event was organized by the Digital Humanities Munich (dhmuc) association and focused on topics including automated data mining for protest events, machine learning, natural language processing, and visualization in digital humanities.

Thinking aloud is a proven method for exploring cognitive processes. It is popular in many disciplines, also for uncovering user experience during the interaction with a computer interface. After some background knowledge on thinking aloud, Thinkie is proposed. It puts thinking aloud onto a mobile platform made of iPhone and iPad. An iPhone with Siri speech recognition (http://www.apple.com/de/ios/siri/) serves as extended microphone. It records the thinking aloud protocol and Siri transcribes it. In parallel the session can be videotaped on the iPad or a different device. An iPad screen integrates the session video and the Siri transcript text, so that the researcher can rework (synchronize, model) it. The transcript is transported via iCloud, the movie via iTunes synchronization. The app is in a proof-of-concept state. Thinkie still falls short in some points of practical impact. A main issue is that the Siri takes are still too short. This may change in the near future. [ABSTRACT FROM AUTHOR]