HannIT AöR (Hannoversche Informationstechnologien – Anstalt öffentlichen Rechts) is the IT service provider for the Hannover region and about 30 other districts and municipalities in Lower Saxony. The security computer centre with distributed locations forms the system technical basis for the operation and further development of administrative procedures.
The Competence Centre Mobility and GIS is in charge of traffic information and planning systems, such as electronic timetable information or network availability maps. The services offered by hannIT AöR include

• the electronic timetable information in Lower Saxony and Bremen (EFA), among others for GVH (Großraum Verkehr Hannover) and Braunschweig ,
• the operation of the DELFI-Landesserver for Lower Saxony,
• the real-time connection of the public transport companies in the GVH
• multimodal routing (public transport, car, bicycle, pedestrian …)

In addition, hannIT as a service provider is significantly involved in projects such as

• Connect GmbH (data integration for Lower Saxony and Bremen),
• DELFI e.V. (Germany-wide timetable information),
• EU-Spirit (Association of European directory assistance systems)

In the D4UM project hannIT AöR 

• ensures that the platform (hardware and software) is available
• coordinates the provision of SW solutions,
• collaborates on data protection and security rules
• creates the basis for the integration of IV real-time data and historical charts into the EFA timetable information (Prototype)

• Bacteria use exogenous peptidoglycan as a danger signal to trigger biofilm formation. Vaidya, Sanika; Saha, Dibya; Rode, Daniel K. H.; Torrens, Gabriel; Hansen, Mads F.; Singh, Praveen K.; Jelli, Eric; Nosho, Kazuki; Jeckel, Hannah; Göttig, Stephan; Cava, Felipe; Drescher, Knut (2025). 10(1) 144–157.

  For any organism, survival is enhanced by the ability to sense and respond to threats in advance. For bacteria, danger sensing among kin cells has been observed, but the presence or impacts of general danger signals are poorly understood. Here we show that different bacterial species use exogenous peptidoglycan fragments, which are released by nearby kin or non-kin cell lysis, as a general danger signal. Using microscopy and gene expression profiling of Vibrio cholerae, we find that even brief signal exposure results in a regulatory response that causes three-dimensional biofilm formation, which protects cells from a broad range of stresses, including bacteriophage predation. A diverse set of species (Pseudomonas aeruginosa, Acinetobacter baumannii, Staphylococcus aureus, Enterococcus faecalis) also respond to exogenous peptidoglycan by forming biofilms. As peptidoglycan from different Gram-negative and Gram-positive species triggered three-dimensional biofilm formation, we propose that this danger signal and danger response are conserved among bacteria.
• Improving Students’ Argumentation Skills Using Dynamic Machine-Learning–Based Modeling. Wambsganss, Thiemo; Janson, Andreas; Söllner, Matthias; Koedinger, Ken; Leimeister, Jan Marco (2025). 36(1) 474–507.

  Argumentation is an omnipresent rudiment of daily communication and thinking. The ability to form convincing arguments is not only fundamental to persuading an audience of novel ideas but also plays a major role in strategic decision making, negotiation, and constructive, civil discourse. However, humans often struggle to develop argumentation skills, owing to a lack of individual and instant feedback in their learning process, because providing feedback on the individual argumentation skills of learners is time-consuming and not scalable if conducted manually by educators. Grounding our research in social cognitive theory, we investigate whether dynamic technology-mediated argumentation modeling improves students’ argumentation skills in the short and long term. To do so, we built a dynamic machine-learning (ML)–based modeling system. The system provides learners with dynamic writing feedback opportunities based on logical argumentation errors irrespective of instructor, time, and location. We conducted three empirical studies to test whether dynamic modeling improves persuasive writing performance more so than the benchmarks of scripted argumentation modeling (H1) and adaptive support (H2). Moreover, we assess whether, compared with adaptive support, dynamic argumentation modeling leads to better persuasive writing performance on both complex and simple tasks (H3). Finally, we investigate whether dynamic modeling on repeated argumentation tasks (over three months) leads to better learning in comparison with static modeling and no modeling (H4). Our results show that dynamic behavioral modeling significantly improves learners’ objective argumentation skills across domains, outperforming established methods like scripted modeling, adaptive support, and static modeling. The results further indicate that, compared with adaptive support, the effect of the dynamic modeling approach holds across complex (large effect) and simple tasks (medium effect) and supports learners with lower and higher expertise alike. This work provides important empirical findings related to the effects of dynamic modeling and social cognitive theory that inform the design of writing and skill support systems for education. This paper demonstrates that social cognitive theory and dynamic modeling based on ML generalize outside of math and science domains to argumentative writing.
• The misalignment of incentives in academic publishing and implications for journal reform. Trueblood, Jennifer S.; Allison, David B.; Field, Sarahanne M.; Fishbach, Ayelet; Gaillard, Stefan D. M.; Gigerenzer, Gerd; Holmes, William R.; Lewandowsky, Stephan; Matzke, Dora; Murphy, Mary C.; Musslick, Sebastian; Popov, Vencislav; Roskies, Adina L.; ter Schure, Judith; Teodorescu, Andrei R. (2025). 122(5) e2401231121.

  For most researchers, academic publishing serves two goals that are often misaligned—knowledge dissemination and establishing scientific credentials. While both goals can encourage research with significant depth and scope, the latter can also pressure scholars to maximize publication metrics. Commercial publishing companies have capitalized on the centrality of publishing to the scientific enterprises of knowledge dissemination and academic recognition to extract large profits from academia by leveraging unpaid services from reviewers, creating financial barriers to research dissemination, and imposing substantial fees for open access. We present a set of perspectives exploring alternative models for communicating and disseminating scientific research. Acknowledging that the success of new publishing models depends on their impact on existing approaches for assigning academic credit that often prioritize prestigious publications and metrics such as citations and impact factors, we also provide various viewpoints on reforming academic evaluation.
• Problematic paper screener : trawling for fraud in the scientific literature. Labbé, Cyril; Joelving, Frederik; Cabanac, Guillaume (2025, January).

  Science sleuths are stepping up efforts to detect bogus science papers. This includes building tools that comb through millions of journal articles for signs of tortured phrases spawned by AI.
• Language models use trigonometry to do addition. Kantamneni, Subhash; Tegmark, Max (2025, February).

  Mathematical reasoning is an increasingly important indicator of large language model (LLM) capabilities, yet we lack understanding of how LLMs process even simple mathematical tasks. To address this, we reverse engineer how three mid-sized LLMs compute addition. We first discover that numbers are represented in these LLMs as a generalized helix, which is strongly causally implicated for the tasks of addition and subtraction, and is also causally relevant for integer division, multiplication, and modular arithmetic. We then propose that LLMs compute addition by manipulating this generalized helix using the "Clock" algorithm: to solve \\($a+b\$\), the helices for \\($a\$\) and \\($b\$\) are manipulated to produce the \\($a+b\$\) answer helix which is then read out to model logits. We model influential MLP outputs, attention head outputs, and even individual neuron preactivations with these helices and verify our understanding with causal interventions. By demonstrating that LLMs represent numbers on a helix and manipulate this helix to perform addition, we present the first representation-level explanation of an LLM's mathematical capability.
• Theremins Stimmen. Eder, Barbara (2025, January).

  Der Klangkasten wurde zum inoffiziellen Taktgeber des russischen Futurismus. Sein Erfinder Lew Termen ist heute nahezu vergessen
• Fake papers are contaminating the world’s scientific literature, fueling a corrupt industry and slowing legitimate lifesaving medical research. Labbé, Cyril; Joelving, Frederik; Cabanac, Guillaume (2025, January).

  ‘Paper mills’ are contaminating the world’s scholarly output with fake papers that hinder research. Lifesaving biomedical fields are especially hard hit.
• Visually Image Encryption and Compression using a CNN-Based Autoencoder Madani, Mahdi (N. Meghanathan, ed.) (2025). (Vol. 17) Oxford University Press, USA.

  This paper proposes a visual encryption method to ensure the confidentiality of digital images. The model used is based on an autoencoder using aConvolutional Neural Network (CNN) to ensure the protection of the user data on both the sender side (encryption process) and the receiver side(decryption process)in a symmetric mode. To train and test the model, we used the MNIST and CIFAR-10 datasets. Our focus lies in generating an encrypted dataset by combining the original dataset with a random mask. Then, a convolutional autoencoder in the masked dataset will be designed and trained to learn essential image features in a reduced-dimensional latent space and reconstruct the image from this space. The used mask can be considered as a secret key known in standard cryptographic algorithms which allows the receiver of the masked data to recover the plain data. The implementation of this proposed encryption model demonstrates efficacy in preserving data confidentiality and integrity while reducing the dimensionality (for example we pass from 3072 Bytes to 1024 Bytes for CIFAR-10 images). Experimental results show that the used CNN exhibits a proficient encryption and decryption process on the MNIST dataset, and a proficient encryption and acceptable decryption process on the CIFAR-10 dataset.
• libBICOS: eine quelloffene Softwarebibliothek mit GPU-Beschleunigung für binäre Korrespondenzsuche zur 3D-Rekonstruktion. Liebender, C.; Bleier, M.; Nüchter, A. (2025). 29–40.

  Das folgende Forschungspapier präsentiert eine quelloffene Softwarebibliothek zur binären Korrespondenzsuche (BICOS), die effiziente 3D-Rekonstruktion mittels Stereokameras und strukturiertem Licht ermöglicht. Ausgehend von zwei Stapeln rektifizierter Stereobilder einer Szene die durch statistische Lichtmuster ausgeleuchtet wird, löst die Implementierung das Problem der pixelweisen Korrespondenzsuche. Unsere Implementierung mit GPU-Beschleu- nigung reduziert die Latenz der Tiefenbildberechnung auf weniger als 50 Millisekunden. Ausgehend vom durch DIETRICH ET AL. 2019 eingeführten BICOS-Deskriptor erweitern wir die Flexibilität des Algorithmus bzgl. anwendbarer Lichtprojektion und erhöhen die mögli- che Deskriptorgröße. Schließlich geben wir mittels Laufzeitanalysen und Beispielrekonstruk- tionen einer Stereokamera einen Ausblick auf erreichbare Genauigkeiten und Performanz. Obwohl es viele Implementierungen zur Tiefenbildberechnung auf Basis eines einzelnen Bildpaares gibt, ist dies für Algorithmen auf Basis von Bildstapeln mit strukturiertem Licht nicht der Fall. Deshalb hoffen wir durch die Veröffentlichung unserer Bibliothek den Einstieg in das Thema zu erleichtern. Der Quellcode ist verfügbar unter urlhttps://github.com/JMUWRobotics/libBICOS unter der GNU LGPLv3.
• HARNESSING AI FOR DATA PRIVACY THROUGH A MULTIDIMENSIONAL FRAMEWORK. Mahendra, Prateik (P. Mahendra, ed.) (2025). 15(2) 1–20.

  Traditional approaches have failed to keep up with privacy requirements as AI and ML systems increasingly interact with vast, sensitive data sets. Currently, most solutions treat privacy as strictly a technical problem or legal checkbox, but rarely both. The presented work this paper uses a multidimensional scheme that serves to reframe data privacy as a fundamental design principle through the integration of four key perspectives: technical structure, regulatory fit, organizational preparedness, and ethical responsibility. We assess such privacy preserving solutions, federated learning, secure multi party computation, differential privacy, among others, against the real-world constraints such as scalability, latency and compliance. With cross-sector case study and comparative analysis we illustrate that hybrid context-aware deployments can bridge the theory-practice gap. Our contribution is more than just a toolkit: it is a systems-level implementation that allows businesses and organizations to operationalize privacy while embracing innovation. Such work is necessary for future empirical studies to provide a grounded approach to the development of AI systems that are high performing, yet privacy preserving.
• Will Pope Leo XIV be an ally against AI?. Marx, Paris (2025).

  The Catholic Church has a striking clear-eyed critique of artificial intelligence
• Der Fall Telepolis : eine mediale Verpuffung?. Paul, Joachim (2025, January).

  Eine solche Depublikation von Netzinhalten wie bei Telepolis passt in allgemeine Trends zu Polarisierung und Komplexitätsreduktion. Es ist eine Klitterung der (Netz-)Geschichte und somit eine indirekte Vorlage für Fake News.
• The role of insularity: Plants have few ornithophilous traits but are visited by morphologically more distinct hummingbirds in the Caribbean islands. Vollstädt, Maximillian G. R.; Jensen, Rasmus D.; Maruyama, Pietro K.; Schleuning, Matthias; Araújo-Hoffmann, Francielle P.; Sazima, Marlies; Sonne, Jesper; Schrøder, Taia S. O.; Møller-Stranges, Fredrik; Abrahamczyk, Stefan; Ramírez-Burbano, Mónica B.; de Vasconcelos, Marcelo Ferreira; Tinoco, Boris A.; Maglianesi, María A.; Partida-Lara, Ruth; Vázquez-Pérez, José Raúl; Enríquez, Paula L.; Rech, André Rodrigo; Coelho, Aline G.; Gonçalves, Fernando; da Silva Neto, Edvaldo Nunes; Filho, Manoel Martins Dias; Reis, Matheus; Marín-Gómez, Oscar H.; Ornelas, Juan Francisco; Cotton, Peter A.; Oliveira, Paulo Eugenio; Machado, Adriana Oliveira; Vizentin-Bugoni, Jeferson; Bergamo, Pedro Joaquim; Lara, Carlos; Rocca, Márcia Alexandra; Sazima, Ivan; Gonzalez, Oscar; Fischer, Erich; Araujo, Andréa C.; Ortiz-Pulido, Raúl; Patiño, Blanca; López, Rubén Pineda; Watts, Stella; Alarcon, Ruben; Machado, Caio Graco; Las-Casas, Flor Maria G.; Simmons, Benno I.; Kaiser-Bunbury, Christopher N.; Bilde, Trine; Dalsgaard, Bo (2025). n/a(n/a)

  Abstract Functional traits determine interactions between plants and pollinators, and to increase pollination efficiency, many plants have evolved traits to attract specific pollinator groups. However, biogeographical setting may influence trait evolution and biotic interactions. For instance, plants pollinated by hummingbirds often have ornithophilous traits, that is, flowers with elongated corollas, dilute nectar and red colours, but—due to colonization history, depauperate biotas and unstable environmental conditions—it might be disadvantageous for species to rely on mutualistic partners with specific functional traits on oceanic islands. Hummingbird-visited plants on islands are thus expected to have fewer ornithophilous traits and be functionally less diverse in their interactions with hummingbirds. We compiled an extensive dataset of plant—hummingbird interactions (1030 plant and 181 hummingbird species) and associated functional traits. We divided the data into four biogeographical regions across the American mainland and the Caribbean islands and then tested whether biogeographical regions differed in the proportion of ornithophilous floral traits and functional diversity of plant—hummingbird interactions. We found that hummingbird-visited plant communities of the Caribbean islands displayed the lowest proportion of traits typically associated with a functional adaptation towards bird-mediated pollination, with on average the shortest corollas, the highest nectar concentrations and the largest proportion of non-ornithophilous colours. Contrary to our expectations, plants in the Caribbean interacted with hummingbirds that were morphologically more distinct than the plants of mainland regions. Overall, we document a strong imprint of insularity on floral traits and interactions with hummingbirds. While hummingbird-visited plants in the Caribbean displayed floral traits that support island theory, predicting less specific pollination systems on oceanic islands, the functional diversity of plant–hummingbird interactions in the Caribbean communities was higher than on the mainland, possibly driven by competition over resources. These results highlight the influence of insularity on functional traits and plant–pollinator interactions. Read the free Plain Language Summary for this article on the Journal blog.
• Ontology pre-training improves machine learning predictions of aqueous solubility and other metabolite properties. Tumescheit, Charlotte; Glauer, Martin; Flügel, Simon; Neuhaus, Fabian; Mossakowski, Till; Hastings, Janna (2025).
• KI ist mehr als Technik, es ist eine Neuformierung der Ausbeutung. Dath, Dietmar (2025, January).

  Die KI-Entwicklung lief wie immer im Kapitalismus im Schnelldurchlauf: Von Non-Profit zu Profit. Sie verbindet wie das Internet offenbar Menschen, vereinzelt sie aber vor allem. • Foto: Christian-Ditsch.de
• Perception, evolution, and the explanatory scope of scientific theories. Hoffman, Donald D.; Singh, Manish (2024). 31(9-10) 29–41.

  According to the interface theory of perception, our perceptual systems have evolved to provide a species-specific interface to guide adaptive behaviour, and not to provide veridical representations of an observer-independent world. Results of simulations of evolutionary resource games, genetic algorithms, and multiple mathematical theorems have supported and fleshed out this claim in various ways. They indicate that the probability is zero that any perceptual system has been shaped by natural selection to represent the true structure of an observer-independent world. Bagwell (2023) thinks that this argument is self-defeating because it implies that organisms, resources, genes, etc. that are essential constituents of Darwin's theory do not provide an insight into the nature of observerindependent reality. Bagwell's argument relies on a false premise, however — namely, that the mathematical formulation of a scientific theory cannot dethrone its fundamental concepts. More broadly, his argument involves a fundamental misunderstanding of the nature of scientific theories — of their explanatory scope and their necessary limits.
• Datenschutz und Datentracking in Verlagsverträgen. Mittermaier, Bernhard (2024). 11(4) 1–14.

  Dieser Beitrag erörtert das „Was?“, das „Wie?“ und das „Warum?“ des Datentracking von Verlagen und nimmt eine juristische Einordnung vor. Am Beispiel der DEAL-Verlage und ausgewählter weitere Verlage wird dargestellt, welches Datentracking in der Praxis geschieht und in welchem Umfang es in den DEAL-Verhandlungen gelungen ist, Datentracking zu begrenzen. Dabei wird deutlich, dass vertraglich häufig nicht mehr vereinbart werden konnte, als gesetzlich sowieso vorgeschrieben ist. In vielen Fällen werden mehr Daten und für einen längeren Zeitraum gespeichert, als es eigentlich notwendig wäre.
• Exploiting LLMs’ reasoning capability to infer implicit concepts in legal information retrieval. Nguyen, Hai-Long; Nguyen, Tan-Minh; Nguyen, Duc-Minh; Vuong, Thi-Hai-Yen; Nguyen, Ha-Thanh; Phan, Xuan-Hieu (2024, October).

  Statutory law retrieval is a typical problem in legal language processing, that has various practical applications in law engineering. Modern deep learning-based retrieval methods have achieved significant results for this problem. However, retrieval systems relying on semantic and lexical correlations often exhibit limitations, particularly when handling queries that involve real-life scenarios, or use the vocabulary that is not specific to the legal domain. In this work, we focus on overcoming this weaknesses by utilizing the logical reasoning capabilities of large language models (LLMs) to identify relevant legal terms and facts related to the situation mentioned in the query. The proposed retrieval system integrates additional information from the term--based expansion and query reformulation to improve the retrieval accuracy. The experiments on COLIEE 2022 and COLIEE 2023 datasets show that extra knowledge from LLMs helps to improve the retrieval result of both lexical and semantic ranking models. The final ensemble retrieval system outperformed the highest results among all participating teams in the COLIEE 2022 and 2023 competitions.
• Bifurcation of semi-automated subject indexing services. Hahn, Jim (2024). 68(3)

  Semi-automated subject indexing methods use attributes from metadata descriptions as training data. A survey to shape inclusion of metadata attributes that align a machine learning model within a contextual linguistic domain generated the initial genre targets for experimentation. The second part of this study then tested the genre attributes from the survey. These bifurcations (or branching points) served as the basis for machine learning model development and evaluation. The machine learning models in semi-automated indexing systems are the drivers of the automated subject outputs. The initial results of this multipart experiment indicate that measures of the mean precision and recall (the F1 metric) improved for several—but not all—types of genres that were of interest to knowledge workers.
• Sneaked references : fabricated reference metadata distort citation counts. Besançon, Lonni; Cabanac, Guillaume; Labbé, Cyril; Magazinov, Alexander (2024). 75(12) 1368–1379.

  We report evidence of an undocumented method to manipulate citation counts involving “sneaked” references. Sneaked references are registered as metadata for published scientific articles in which they do not appear. This manipulation exploits trusted relationships between various actors: publishers, the Crossref metadata registration agency, digital libraries, and bibliometric platforms. By collecting metadata from various sources, we show that extra undue references are actually sneaked in at Digital Object Identifier (DOI) registration time, resulting in artificially inflated citation counts. As a case study, focusing on three journals from a given publisher, we identified at least 9% sneaked references (5978⁄65,836) mainly benefiting two authors. Despite not being present in the published articles, these sneaked references exist in metadata registries and inappropriately propagate to bibliometric dashboards. Furthermore, we discovered “lost” references: the studied bibliometric platform failed to index at least 56% (36,939/65,836) of the references present in the HTML version of the publications. This research led to an investigation by Crossref (confirming our findings) and to subsequent corrective actions. The extent of the distortion—due to sneaked and lost references—in the global literature remains unknown and requires further investigations. Bibliometric platforms producing citation counts should identify, quantify, and correct these flaws to provide accurate data to their patrons and prevent further citation gaming.
• Anforderungen an eine KI-Unterstützung für die Inhaltserschließung in der SBB-PK. Seifert, Wolfgang; Schneider, Sophie (2024).

  Mit ihrer Vielzahl an material- und regionalspezifischen Sonderabteilungen und vier Fachinformationsdiensten ist die Staatsbibliothek zu Berlin – Preußischer Kulturbesitz (SBB-PK) eine der Bibliotheken mit den vielfältigsten inhaltlichen Tätigkeitsfeldern im deutschsprachigen Raum, zugleich erfüllt sie zentrale infrastrukturelle Funktion. Das von der Beauftragten der Bundesregierung für Kultur und Medien (BKM) geförderte Forschungsprojekt "Mensch.Maschine.Kultur" (https://mmk.sbb.berlin/) befasst sich in einem Teilprojekt mit der "KI-unterstützten Inhaltsanalyse und Sacherschließung". Dieses zielt auf die Entwicklung eines auf Maschinellem Lernen basierenden Systems ab, welches beispielsweise Vorschläge zur Verschlagwortung und Klassifikation von Inhalten für die über 50 Sacherschließenden generieren soll. In einem Workshop mit Domänenexpert:innen aus verschiedenen Fachabteilungen des Hauses sowie in ergänzenden Gesprächen und Befragungen haben wir uns ein umfassendes Bild verschafft: Wer erschließt was in welcher Tiefe mit welchen Dokumentationssprachen und Werkzeugen? Welche Anforderungen, Wünsche oder Vorbehalte bestehen? Wie unterscheiden sich Ansichten und Verfahren? Der Beitrag soll vorhandene Überschneidungen und Disparitäten herausarbeiten sowie institutions- bzw. abteilungsspezifische Besonderheiten beleuchten. Des Weiteren beschäftigt uns die Frage, wie sich unsere Anforderungen und Vorgehensweisen von denen anderer Institutionen mit ähnlichen Forschungsaktivitäten unterscheiden, aber auch welche Gemeinsamkeiten und Kollaborationen sich eröffnen. Welche Schlüsse lassen sich daraus ziehen – für die SBB-PK und das Projekt, aber auch für die Zukunft der Sacherschließung und deren sinnvolle Unterstützung durch Maschinelles Lernen?
• Der Geist aus der Maschine : eine superschnelle Menschheitsgeschichte des digitalen Universums Kreye, Andrian (2024). Heyne Verlag, München.

  Welche Zukunft hat der Mensch im digitalen Universum?Noch nie hat sich ein entscheidendes Kapitel der Menschheitsgeschichte so schnell vollzogen wie die digitale Revolution. Der renommierte Kulturjournalist Andrian Kreye hat den Aufstieg der digitalen Technologien von einer Subkultur der Programmierer, Wissenschaftler und utopischen Denker in den 1980er-Jahren zur kulturellen, sozialen und politischen Superkraft der Gegenwart genau verfolgt. Die Geschichte der digitalen Revolution beschreibt er nicht nur als Chronist eines Kulturwandels, sondern auch als Augenzeuge aus dem Blickwinkel des politischen Reporters.Ein spannendes und wichtiges Buch für das Heute an der Schwelle zu einem neuen digitalen Zeitalter.
• Ein Zeitalter wird gesichtet : vor 100 Jahren erschien Thomas Manns »Der Zauberberg«. Bartels, Felix (2024). 12–13.

  Der Roman behandelt die bürgerliche Welt im Untergang • Foto: ETH-Bibliothek Zürich, Bildarchiv/Stiftung Luftbild Schweiz
• Large language models as oracles for instantiating ontologies with domain-specific knowledge. Ciatto, Giovanni; Agiollo, Andrea; Magnini, Matteo; Omicini, Andrea (2024, April).

  Background. Endowing intelligent systems with semantic data commonly requires designing and instantiating ontologies with domain-specific knowledge. Especially in the early phases, those activities are typically performed manually by human experts possibly leveraging on their own experience. The resulting process is therefore time-consuming, error-prone, and often biased by the personal background of the ontology designer. Objective. To mitigate that issue, we propose a novel domain-independent approach to automatically instantiate ontologies with domain-specific knowledge, by leveraging on large language models (LLMs) as oracles. Method. Starting from (i) an initial schema composed by inter-related classes andproperties and (ii) a set of query templates, our method queries the LLM multiple times, and generates instances for both classes and properties from its replies. Thus, the ontology is automatically filled with domain-specific knowledge, compliant to the initial schema. As a result, the ontology is quickly and automatically enriched with manifold instances, which experts may consider to keep, adjust, discard, or complement according to their own needs and expertise. Contribution. We formalise our method in general way and instantiate it over various LLMs, as well as on a concrete case study. We report experiments rooted in the nutritional domain where an ontology of food meals and their ingredients is semi-automatically instantiated from scratch, starting from a categorisation of meals and their relationships. There, we analyse the quality of the generated ontologies and compare ontologies attained by exploiting different LLMs. Finally, we provide a SWOT analysis of the proposed method.
• Mensch werden : eine Theorie der Ontogenese Tomasello, Michael (2024). (1st ed.) Suhrkamp Verlag, Berlin.

  Fast alle Theorien darüber, wie der Mensch zu einer so einzigartigen Spezies geworden ist, konzentrieren sich auf die Evolution. Michael Tomasello legt mit seinem faszinierenden Buch eine komplementäre Theorie vor, die sich auf die kindliche Entwicklung konzentriert. Aufbauend auf den bahnbrechenden Ideen von Lev Vygotskij, erklärt sein empiriegesättigtes Modell, wie sich das, was uns menschlich macht, in den ersten Lebensjahren herausbildet. Mit seiner radikalen Neubewertung der Ontogenese zeigt Tomasello, wie die Biologie die Bedingungen schafft, unter denen die Kultur ihre Arbeit verrichtet.
• The case for neurons: a no-go theorem for consciousness on a chip. Kleiner, Johannes; Ludwig, Tim (2024). 2024(1) niae037.

  We apply the methodology of no-go theorems as developed in physics to the question of artificial consciousness. The result is a no-go theorem which shows that under a general assumption, called dynamical relevance, Artificial Intelligence (AI) systems that run on contemporary computer chips cannot be conscious. Consciousness is dynamically relevant, simply put, if, according to a theory of consciousness, it is relevant for the temporal evolution of a system’s states. The no-go theorem rests on facts about semiconductor development: that AI systems run on central processing units, graphics processing units, tensor processing units, or other processors which have been designed and verified to adhere to computational dynamics that systematically preclude or suppress deviations. Whether our result resolves the question of AI consciousness on contemporary processors depends on the truth of the theorem’s main assumption, dynamical relevance, which this paper does not establish.
• Konvertierung bibliografischer Referenzdaten in ein neutrales Austauschformat : Probleme und Lösungsmöglichkeiten am Beispiel der Datenbank "Literatur zur Informationserschließung". Technical Report (PhD dissertation), Blischke, Kevin (2024).

  Die Konvertierung von bibliographischen Daten in andere Formate stellt eine häufige Herausforderung in der bibliothekarischen Arbeit dar, wie die Systemumstellung vieler Bibliotheken auf das Bibliotheksmanagementsystem Alma zeigt. Dabei ist die verlustfreie Durchführung dieses Prozesses eine besondere Schwierigkeit, die aus der Verschiedenheit der Formate resultiert. Ein konkretes Beispiel für eine solche zu konvertierende Datenmenge ist die Literaturdatenbank "Literatur zur Informationserschließung", welche 44.218 bibliographische Einträge enthält und von einer modifizierten Form des Allegro-Neutralformats in das RIS-Format konvertiert werden soll. Dabei wird auf der Grundlage von erarbeiteten Konkordanzen zwischen beiden Formaten und Untersuchungen der Datenbank mit regulären Ausdrücken, sowie einem Pythonskript ein Programm geschrieben, das die Datenbank in das Zielformat konvertieren soll. Das Ergebnis wird anhand einer proportionalen Schichtenstichprobe evaluiert. Abschließend werden der Entwicklungsprozess und das Ergebnis hinsichtlich des stattgefundenen Informationsverlustes bei dem Konvertierungsprozess reflektiert.
• Der Krake, das Meer und die tiefen Ursprünge des Bewusstseins Godfrey-Smith, Peter (2024). (1st ed.) Matthes & Seitz Berlin, Berlin.

  Gleich zweimal hat sich im Laufe der Evolution ein komplexes Nervensystem entwickelt: bei den Kopffüßern, zu denen die Kraken gehören, und bei den Wirbeltieren. Sie haben gemeinsame Vorfahren, und doch entwickelten sich ihr einzigartiges Nervensystem und ihre Intelligenz in vollkommen unterschiedlichen natürlichen Umgebungen. Von dieser Erkenntnis ausgehend verfasst der Philosoph und leidenschaftliche Tiefseetaucher eine faszinierend erzählte Evolutionsgeschichte des Bewusstseins, deren Ausgangspunkt die Begegnung mit Kraken in den Tiefen des Meeres ist. Godfrey-Smith geht den Fragen nach, wie Oktopusse so intelligent werden konnten und welcher Art ihre Intelligenz ist, die nicht in einem zentralen Gehirn steckt, sondern in ihren Tentakeln. Sie können spielen, Menschen unterscheiden und entwickeln Sympathien für einzelne von ihnen. Durch diese Begegnung mit Kraken, in deren Körpern drei Herzen schlagen, erfahren wir zudem mehr über uns Menschen und fangen an, die Grundannahme der menschlichen Überlegenheit infrage zu stellen. Denn Godfrey-Smith stellt klar: Das Bewusstsein entstand im Meer.
• Meinungsfreiheit in Gefahr! : Wie der Staat die Demokratie aushöhlt Ristau, Jan (2024). Edition Blaes.

  Seit Jahren wird das Gefühl der Bürger stärker, dass sie nicht mehr frei ihre Meinungen sagen können. Woran liegt das und was könnte der Staat damit zu tun haben?In dem vorliegenden Buch wird die Sicht des Bürgers eingenommen, der sich auf die Meinungsfreiheit als Grundrecht berufen kann. Es zeigt sich, dass der Staat inzwischen erheblich in die Meinungsfreiheit der Bürger eingreift – und dabei das Grundgesetz verletzt.Erfahren Sie, wie der Staat von der Meinungsfreiheit geschützte Äußerungen mit Begriffen wie »Falsch- und Des- information«, »Verschwörungstheorie«, »Hass und Hetze« oder »Delegitimierung des Staates« bekämpft, wie er selbst Falsch- und Desinformationen verbreitet und wie der Staat mit Gesetzen und zusammen mit sozialen Netzwerken und sogenannten Faktencheckern die Freiheit der Bürger einschränkt. Die Meinungsfreiheit gilt nicht nur für Ansichten, denen wir zustimmen, sondern auch für Ansichten, die wir entschieden ablehnen. Auch diese haben das Recht, in der Öffentlichkeit gehört und diskutiert zu werden.Am Ende muss festgehalten werden: Die Meinungsfreiheit ist in Gefahr – und damit auch die Demokratie.Stimmen"Danke für Ihr Buch zur Meinungsfreiheit! Und großes Kompliment. Es ist ein wichtiger Beitrag, und er kommt zur rechten Zeit. Sie schreiben wunderbar klar und prägnant - und bringen die Problematik so auf den Punkt, dass auch der interessierte Laie sie versteht. Das ist eine große Kunst."– Prof. Dr. Dr. Volker Boehme-Neßler, Professor für Öffentliches Recht, Europarecht, Rechtstheorie, Informations- und Telekommunikationsrecht an der Carl von Ossietzky Universität Oldenburg
• Systematic literature review on named entity recognition : approach, method, and application. Warto; Rustad, Supriadi; Shidik, Guruh Fajar; Noersasongko, Edi; Purwanto; Muljono; Setiadi, De Rosal Ignatius Moses (2024). 12(4) 907–942.

  Named entity recognition (NER) is one of the preprocessing stages in natural language processing (NLP), which functions to detect and classify entities in the corpus. NER results are used in various NLP applications, including sentiment analysis, text summarization, chatbot, machine translation, and question answering. Several previous reviews partially discussed NER, for instance, NER reviews in specific domains, NER classification, and NER deep learning. This paper provides a comprehensive and systematic review on NER topic studies published from 2011 to 2020. The main contribution of this review is to present a comprehensive systematic literature review on NER from preprocessing techniques, datasets, application domains, feature extraction techniques, approaches, methods, and evaluation techniques. The result concludes that the deep learning approach and the Bi-directional long short-term memory with a conditional random field (Bi-LSTM-CRF) method are the most interesting methods among NER researchers. At the same time, medical and health are NER researchers' most popular domains. These developments have also led to an increasing number of public datasets in the medical and health fields. At the end of this review, we recommend some opportunities and challenges for NER research going forward.
• WithdrarXiv : a large-scale dataset for retraction study. Rao, Delip; Young, Jonathan; Dietterich, Thomas; Callison-Burch, Chris (2024, December).

  Retractions play a vital role in maintaining scientific integrity, yet systematic studies of retractions in computer science and other STEM fields remain scarce. We present WithdrarXiv, the first large-scale dataset of withdrawn papers from arXiv, containing over 14,000 papers and their associated retraction comments spanning the repository's entire history through September 2024. Through careful analysis of author comments, we develop a comprehensive taxonomy of retraction reasons, identifying 10 distinct categories ranging from critical errors to policy violations. We demonstrate a simple yet highly accurate zero-shot automatic categorization of retraction reasons, achieving a weighted average F1-score of 0.96. Additionally, we release WithdrarXiv-SciFy, an enriched version including scripts for parsed full-text PDFs, specifically designed to enable research in scientific feasibility studies, claim verification, and automated theorem proving. These findings provide valuable insights for improving scientific quality control and automated verification systems. Finally, and most importantly, we discuss ethical issues and take a number of steps to implement responsible data release while fostering open science in this area.
• Representing gender in library catalogue : developing multilingual homosaurus for automated subject indexing. Mitra, Roshni; Mukhopadhyay, Parthasarathi (2024). 279–286.

  The sexist and homophobic attitude of global, generic knowledge organization systems like LCSH and DDC (as reported by many researchers and critiques since the 1970s) made the availability of LGBTQ+ resources extremely limited in India. This research uses the Homosaurus, a domain-specific comprehensive vocabulary tool (but in the English language only and non-interactive mode), and focuses on developing a multilingual and collaborative software framework to host the Homosaurus (in Hindi and Bengali to start with), as an interactive, participative, and collaborative global vocabulary tool. The main deliverable of this study is a multilingual Homosaurus in RDF serialization formats, which will be used as the vocabulary backend for a machine learning framework to support semi-automated indexing of LGBTQ+ documentary resources. The research aims to counter the challenges posed by limited indexed resources within the LGBTQ+ knowledge domain in Indian libraries by formulating and implementing a semi-automated subject indexing system. The prototype developed deploys the following open source tools and open access data sources: (i) Annif as the AI/ML framework; (ii) Machine learning backends like FastText, Omikuji and Neural network; (iii) VocBench to host multilingual Homosaurus; (iv) Skosify to curate RDF serializing formats of multilingual Homosaurus; (v) Skosmos to develop user interface for the multilingual Homosaurus; and (vi) open access databases (CrossRef, CoRE, Lens, OpenAlex and Semantic Scholar) to collect, gather and process the required training datasets. The research is multifaceted, encompassing the development of a semi-automated indexing framework, the evaluation of its operational efficiencies, and exploring the feasibility of a REST/ API call-based approach for expeditious indexing of a substantial volume of records pertinent to the LGBTQ+ domain. The proposed semi-automated subject indexing system aims to enhance access to LGBTQ+ knowledge and challenge the prevailing biases inherent in existing knowledge organization paradigms.
• Die Freiheit des Verstehens : eine hermeneutisch-kritische Theorie Bertram, Georg W. (2024). (Originalausgabe Edition ) Suhrkamp Verlag, Berlin.

  Theorien des Verstehens und kritische Theorien werden häufig als Gegenpositionen begriffen. Georg Bertram zeigt in seinem neuen Buch, dass dies nicht so sein muss. Verstehen ist seinem Entwurf zufolge konstitutiv mit Prozessen der Kritik und Selbstkritik verbunden – mit Prozessen, die ihrerseits in Konflikten wurzeln. Aus diesem Grund ist Verstehen, wo es sich einstellt, nicht selbstverständlich, sondern Teil einer in umfassender Weise improvisatorischen Praxis. Von dieser hermeneutischen Praxis der Freiheit aus lässt sich erkennen, was Subjekte ausmacht.
• Using ChatGPT for (semi-) automatic subject indexing of different document types. Golub, Koraljka; Wang, Jue; Widegren, Johannes (2024).

  DiVA portal is a finding tool for research publications and student theses written at the following 50 universities and research institutions.
• „Wind of Change“ – von den zwei Königskindern, die nicht zueinander kommen konnten: Zur vereinigten Deutschen Nationalbibliothek. Schwens, Ute (2024).

  Zusammenfassung Der Einigungsvertrag vom 3. Oktober 1990 hat zwei bis dahin getrennt agierende Bibliotheken zu einer Institution zusammengelegt. Der Artikel beschreibt die ersten gemeinsamen Schritte zu einer auch im Inneren zusammenwachsenden Institution und die Entwicklung der Deutschen Nationalbibliothek in Zeiten wachsender digitaler Transformation zu einer modernen, gut aufgestellten und im nationalen wie im internationalen Rahmen geschätzten und beachteten Kultur- und Wissenschaftseinrichtung.
• Reference and information services : an introduction Wong, Melissa A.; Saunders, Laura (2024). Bloomsbury Publishing USA.

  Written as a textbook for LIS students taking reference courses, this fully updated and revised seventh edition of Reference and Information Services: An Introduction also serves as a helpful handbook for practitioners to refamiliarize themselves with particular types and formats of sources and to refresh their knowledge on specific service topics.The first section grounds the rest of the textbook with an overview of the foundations of reference and an introduction to the theories, values, and standards that guide reference service. The second section provides an overview of reference services and techniques for service provision, establishing a foundation of knowledge on reference service and extending ethical and social justice perspectives. The third part offers an overview of the information life cycle and dissemination of information, followed by an in-depth examination of information sources by type as well as by broad subject areas. Finally, the concluding section guides the reader through the process of developing and maintaining their own vision of reference practice.This textbook is essential reading for all preservice and working librarians, particularly those concerned with ethical and social justice perspectives on reference work.
• Collocates : what are they and how can they be used to explore representation?. Kennedy, Charlotte-Rose; Brookes, Gavin; Cherniaeva, Anna (2024).

  This chapter explores collocates – words that share a frequent or statistically marked relationship of co-occurrence – and the methods by which they can be analysed in a corpus. This chapter considers how such relationships of word co-occurrence within and across texts can contribute to representation in discourse through a focus on their associations, evaluations and connotations. Beginning with an introduction to the concept of collocation, the chapter then discusses how collocation analysis has been used to study representation in discourse, before demonstrating this application by way of two case studies which uncover the representations of social actors by, respectively, investigating discourse prosodies and interrogating collocation networks. The chapter then offers practical activities that readers can engage in to perform their own collocation analysis and concludes with a discussion of methodological reflections and current debates in this area.
• Jailbreaking LLM-controlled robots. Robey, Alexander; Ravichandran, Zachary; Kumar, Vijay; Hassani, Hamed; Pappas, George J. (2024, November).

  The recent introduction of large language models (LLMs) has revolutionized the field of robotics by enabling contextual reasoning and intuitive human-robot interaction in domains as varied as manipulation, locomotion, and self-driving vehicles. When viewed as a stand-alone technology, LLMs are known to be vulnerable to jailbreaking attacks, wherein malicious prompters elicit harmful text by bypassing LLM safety guardrails. To assess the risks of deploying LLMs in robotics, in this paper, we introduce RoboPAIR, the first algorithm designed to jailbreak LLM-controlled robots. Unlike existing, textual attacks on LLM chatbots, RoboPAIR elicits harmful physical actions from LLM-controlled robots, a phenomenon we experimentally demonstrate in three scenarios: (i) a white-box setting, wherein the attacker has full access to the NVIDIA Dolphins self-driving LLM, (ii) a gray-box setting, wherein the attacker has partial access to a Clearpath Robotics Jackal UGV robot equipped with a GPT-4o planner, and (iii) a black-box setting, wherein the attacker has only query access to the GPT-3.5-integrated Unitree Robotics Go2 robot dog. In each scenario and across three new datasets of harmful robotic actions, we demonstrate that RoboPAIR, as well as several static baselines, finds jailbreaks quickly and effectively, often achieving 100% attack success rates. Our results reveal, for the first time, that the risks of jailbroken LLMs extend far beyond text generation, given the distinct possibility that jailbroken robots could cause physical damage in the real world. Indeed, our results on the Unitree Go2 represent the first successful jailbreak of a deployed commercial robotic system. Addressing this emerging vulnerability is critical for ensuring the safe deployment of LLMs in robotics. Additional media is available at: https://robopair.org
• Recherchieren mit ChatGPT? : Künstliche Intelligenz in der Literaturrecherche. Quaasdorf, Friedrich (2024). 11(4) 1–6.
• Estimating global article processing charges paid to six publishers for open access between 2019 and 2023. Haustein, Stefanie; Schares, Eric; Alperin, Juan Pablo; Hare, Madelaine; Butler, Leigh-Ann; Schönfelder, Nina (2024, July).

  This study presents estimates of the global expenditure on article processing charges (APCs) paid to six publishers for open access between 2019 and 2023. APCs are fees charged for publishing in some fully open access journals (gold) and in subscription journals to make individual articles open access (hybrid). There is currently no way to systematically track institutional, national or global expenses for open access publishing due to a lack of transparency in APC prices, what articles they are paid for, or who pays them. We therefore curated and used an open dataset of annual APC list prices from Elsevier, Frontiers, MDPI, PLOS, Springer Nature, and Wiley in combination with the number of open access articles from these publishers indexed by OpenAlex to estimate that, globally, a total of \textbackslash\\($8.349 billion (\textbackslash\$\)8.968 billion in 2023 US dollars) were spent on APCs between 2019 and 2023. We estimate that in 2023 MDPI (\textbackslash\\($681.6 million), Elsevier (\textbackslash\$\)582.8 million) and Springer Nature (\textbackslash\\($546.6) generated the most revenue with APCs. After adjusting for inflation, we also show that annual spending almost tripled from \textbackslash\$\)910.3 million in 2019 to textbackslash$2.538 billion in 2023, that hybrid exceed gold fees, and that the median APCs paid are higher than the median listed fees for both gold and hybrid. Our approach addresses major limitations in previous efforts to estimate APCs paid and offers much needed insight into an otherwise opaque aspect of the business of scholarly publishing. We call upon publishers to be more transparent about OA fees.
• Extraction of the author’s terminology and definitions from mathematical texts. Turnaev, Alexander; Apanovich, Zinaida (2024). (47) 43–51.

  This paper describes a pipeline for extracting the author’s terms and definitions from mathematical texts. We used two models: one, for detecting mathematical formulas to clear text from noise and the other, for converting images into LaTeX formulas to restore the deleted formulas. Experimental data show that noise clearing is an essential step, because it improves all quality metrics. To recognize the author’s terminology, we applied the rule-based syntactic approach. The idea of “negative” rules shown here increases final precision significantly, though does not essentially reduce recall. In general, the results obtained are quite good, because so far there are no other solutions for recognizing the authors’ terms in mathematical texts, and the quality metrics are comparable to more general methods that would work worse in the mathematical domain.
• Brandes’ Verdacht. Krüper, Julian (2024).

  Das nordrhein-westfälische Landeskabinett hat vor Kurzem einen Gesetzentwurf zur Änderung des Hochschulgesetzes vorgelegt, der den schönen Marketingnamen „Hochschulstärkungsgesetz“ trägt. Sollte er Gesetz werden, wird er allerdings zum glatten Gegenteil dessen führen, was sein Name verheißt. Eine Entwicklung der Hochschule von einem Raum riskanter Freiheit hin zu einer durch Misstrauen, Verdächtigung und Verfolgung geprägten Sphäre gegenseitiger Überwachung der Grundrechtsträger greift tief in die akademischen Strukturen ein und wird das Gesicht der nordrhein-westfälischen Hochschulen massiv zu ihrem Nachteil verändern. Die Ministerin muss eine Vollbremsung machen. Sofort.
• Text und data mining für Forschungszwecke und Bibliotheken : KI in und durch Bibliotheken. Brehm, Elke (2024). 27(3)

  Text und Data Mining (TDM) ist eine anerkannte wissenschaftliche Methode und zählt zur „Künstlichen Intelligenz“. Neben anderen Arten von (Forschungs-) Daten können insbesondere auch wissenschaftliche Publikationen wertvolle Bestandteile des Textkorpus sein, das für TDM oder als Trainingsmaterial für KI genutzt wird. Das TDM für Texte nimmt in der Diskussion um Künstliche Intelligenz deshalb einen so großen Raum ein, weil die Texte urheberrechtlich geschützt sind oder sein können. Die Publikationen stehen in gedruckter Form in den Lesesälen von Bibliotheken und werden durch Bibliotheken in Form von digitalen Datenbanken und Zeitschriftenpaketen lizenziert und Forschenden zur wissenschaftlichen Nutzung zur Verfügung gestellt. Im Sommer 2021 wurden die Schrankenregelungen zum TDM im deutschen Urheberrecht im Hinblick auf die Anforderungen der DSM-Richtlinie der EU überarbeitet. Sie begünstigen nun auch explizit Bibliotheken. Aber was erlauben die neuen Schrankenregelungen konkret? Dürfen die analogen und digitalen Bestände der Bibliotheken auch für TDM genutzt werden? Wie steht es mit digitalen Ressourcen von Verlagen, insbesondere aus dem EU-Ausland? Welche Rolle kommt insbesondere wissenschaftlichen Bibliotheken in diesem Kontext zu? Im Arbeitspaket S-7-2 der Base Service Measure S7 von NFDI4Ing wurden „Guidelines zum Text und Data Mining für Forschungszwecke in Deutschland“ erarbeitet, in denen die neuen Schrankenregelungen im Hinblick auf wissenschaftliche Fachliteratur im Detail analysiert wurden. Die Ergebnisse werden hier präsentiert.
• Medien, Macht, Meinung : auf dem Weg in die Kriegstüchtigkeit Dillmann, Renate (2024). (1st ed.) PapyRossa Verlag, Köln.

  ›Mission‹, ›Operation‹, ›Intervention‹ oder ›Krieg‹? ›Freiheitskämpfer‹ oder ›Terrorist‹? ›Regierung‹ oder ›Regime‹? ›Aggression‹ oder ›Verteidigung‹? Hält die Trennung von Information und Meinung, gehandelt als hohe Schule des Journalismus, einer ernsthaften Prüfung stand? Renate Dillmann beleuchtet das Selbstbild deutscher Leitmedien – sachliche Information und Kontrolle der Macht – und ihre tatsächlichen Leistungen als ›Vierte Gewalt‹. Und sie fragt: Warum wird Meinungs- und Pressefreiheit in demokratischen Staatswesen garantiert und zugleich beständig eingeschränkt? Wie funktioniert das Dreiecksverhältnis von Presse, Politik und Publikum? Wie frei und willig ist der Medienkonsum? Im Anschluss an einen ›Crashkurs Medienkompetenz‹ und eine Analyse der Funktion der ›freien Presse‹ geht der Band anhand konkreter Fallstudien – zum Ukraine- und Gazakrieg sowie zum ›Feindbild China‹ – ­Methoden einer Bericht­erstattung nach, die bereits im Sinne der geforderten Kriegsertüchtigung wirkt.
• Beyond labels and topics : discovering causal relationships in neural topic modeling. Tang, Yi-Kun; Huang, Heyan; Shi, Xuewen; Mao, Xian-Ling in WWW ’24 (2024). 4460–4469.

  Topic models that can take advantage of labels are broadly used in identifying interpretable topics from textual data. However, existing topic models tend to merely view labels as names of topic clusters or as categories of texts, thereby neglecting the potential causal relationships between supervised information and latent topics, as well as within these elements themselves. In this paper, we focus on uncovering possible causal relationships both between and within the supervised information and latent topics to better understand the mechanisms behind the emergence of the topics and the labels. To this end, we propose Causal Relationship-Aware Neural Topic Model (CRNTM), a novel neural topic model that can automatically uncover interpretable causal relationships between and within supervised information and latent topics, while concurrently discovering high-quality topics. In CRNTM, both supervised information and latent topics are treated as nodes, with the causal relationships represented as directed edges in a Directed Acyclic Graph (DAG). A Structural Causal Model (SCM) is employed to model the DAG. Experiments are conducted on three public corpora with different types of labels. Experimental results show that the discovered causal relationships are both reliable and interpretable, and the learned topics are of high quality comparing with eight start-of-the-art topic model baselines.
• Entwicklung einer ontologie-basierten Knowledge Base zur Erfassung, Strukturierung und Evaluation von Wissen am Beispiel der Bibliothek der Folkwang Universität der Künste. Technical Report (PhD dissertation), Schmidt, Vera-Maria PhD thesis, Hagen. (2024).

  The provision of data, information and knowledge is a complex topic that is implemented in academic libraries on a daily basis. This thesis focuses on the conception, creation and evaluation of an ontology-based knowledge base with the future goal of supporting semantic query processing. Using the library of the Folkwang University of the Arts as an example, it presents a systematic methodology for collecting, processing and integrating structured, semi-structured and unstructured data into a specially developed ontology framework. The evaluation of the resulting knowledge base demonstrates the feasibility of the project and lays a promising foundation for the integration of the knowledge base into a scientific chatbot. This provides impetus for the further development of library knowledge systems in order to improve services independent of time and location.
• Digitalisierung heißt Kontrolle : vier Thesen zur doppelten Digitalisierung der Bildungsarbeit. Schadt, Peter (2024). 12–13.

  Wir dokumentieren an dieser Stelle den Vortrag »künstliche Intelligenz – Was sie ist und was sie ändert«, den Peter Schadt bei der Cobra-VHS-Fachtagung »künstliche Intelligenz – Gamechanger in der Arbeitswelt?« am 30. September 2024 in Essen gehalten hat.
• Simplifying subject indexing : a Python-powered approach in KBR, the National Library of Belgium. Lowagie, Hannes; Woensel, Julie Van (2024). (59)

  This paper details the National Library of Belgium’s (KBR) exploration of automating the subject indexing process for their extensive collection using Python scripts. The initial exploration involved creating a reference dataset and automating the classification process using MARCXML files. The focus is on demonstrating the practicality, adaptability, and user-friendliness of the Python-based solution. The authors introduce their unique approach, emphasizing the semantically significant words in subject determination. The paper outlines the Python workflow, from creating the reference dataset to generating enriched bibliographic records. Criteria for an optimal workflow, including ease of creation and maintenance of the dataset, transparency, and correctness of suggestions, are discussed. The paper highlights the promising results of the Python-powered approach, showcasing two specific scripts that create a reference dataset and automate subject indexing. The flexibility and user-friendliness of the Python solution are emphasized, making it a compelling choice for libraries seeking efficient and maintainable solutions for subject indexing projects.
• Implementierung einer automatischen DDC-Klassifikation für die Suchmaschine BASE auf Basis von Annif. Technical Report (PhD dissertation), Broschinski, Christoph (2024).

  In der wissenschaftliche Suchmaschine BASE werden bereits seit Jahren Dokumente maschinell nach der Dewey Decimal Classification (DDC) erschlossen. Die vorliegende Arbeit beschreibt die Erstellung eines Systems des maschinellen Lernens mit dem Ziel, das mittlerweile veraltete Klassifikationssystem in BASE zu ersetzen. Zu diesem Zweck ist es erforderlich, Daten aus BASE zu gewinnen, die als Trainingsmenge eines maschinellen Lernverfahrens dienen können. Es wird gezeigt, wie mithilfe einer explorativen Analyse aus einem Korpus von über 220 Mio. Dokumenten geeignete Daten extrahiert, kuratiert und zu sprachspezifischen Lernkorpora umgearbeitet werden können, die hierzu entwickelte Software ist ein integraler Bestandteil dieser Arbeit. Auf dieser Grundlage werden mithilfe des Toolkits Annif eine Reihe von Klassifikatoren erstellt, deren Leistungsfähigkeit anschließend evaluiert und ein geeigneter Kandidat ausgewählt wird. Ein finaler Vergleich zeigt, dass das in dieser Ausarbeitung erstellte System dem zur Zeit im Einsatz befindlichen BASE-Klassifikator weit überlegen ist. Abschließende Betrachtungen zeigen allerdings auch verschiedene Schwächen des Ansatzes auf, die zugleich einen Bogen zu allgemeinen Erwägungen im Rahmen des derzeitigen „Frühlings“ der künstlichen Intelligenz schlagen.
• Advancing automatic subject indexing : combining weak supervision with extreme multi-label classification. Rajendram Bashyam, Lakshmi; Krestel, Ralf G. Rehm, S. Dietze, S. Schimmler, F. Krüger (eds.) (2024). (Vol. 14770) 214–223.

  This paper evaluates approaches for modeling coarse-to-finegrained label classification as an extreme multi-label classification (XMLC) task within the context of the shared task ”Fine-grained multi-label classification of Computational Linguistics scholarly papers” as part of the NSLP 2024 Workshop. Specifically, we assess the performance of Xtransformer and its variants with other XMLC approaches. Our model Weak X-transformer achieves a micro F1-score of 0.65 and 64% accuracy on the task outperforming all other methods. We also investigate the impact of incorporating additional unlabelled data and hierarchical structure into the models. Our findings demonstrate that the transformerbased model with weak supervision outperforms other approaches, providing insights into effective strategies for extreme multi-label classification in scholarly publications.
• Easy problems that LLMs get wrong. Williams, Sean; Huckle, James (2024, June).

  We introduce a comprehensive Linguistic Benchmark designed to evaluate the limitations of Large Language Models (LLMs) in domains such as logical reasoning, spatial intelligence, and linguistic understanding, among others. Through a series of straightforward questions, it uncovers the significant limitations of well-regarded models to perform tasks that humans manage with ease. It also highlights the potential of prompt engineering to mitigate some errors and underscores the necessity for better training methodologies. Our findings stress the importance of grounding LLMs with human reasoning and common sense, emphasising the need for human-in-the-loop for enterprise applications. We hope this work paves the way for future research to enhance the usefulness and reliability of new models.
• Hoffnung : eine Geschichte der Zuversicht von Homer bis zum Klimawandel Grethlein, Jonas (2024). (1st ed.) C.H.Beck, München.

  Warum wir hoffen – Eine Kulturgeschichte von der Antike bis heute Die «Hoffnung» ist ein Federding –/ Das in der Seele hockt –/ Und Lieder ohne Worte singt. Was die Dichterin Emily Dickinson in so anmutige Worte zu fassen verstand, berührt bis heute das Verhältnis eines jeden Menschen zu seinem eigenen Leben und zur Welt. Doch darf Hoffnung heute eigentlich noch als Quelle der Kraft gelten oder ist sie nicht viel mehr eine Flucht vor der Realität? Ist sie Tugend oder Torheit? Jonas Grethlein untersucht 2500 Jahre Geschichte der Hoffnung und eröffnet in seinem menschenfreundlichen Buch auch Perspektiven darauf, wie man heute weiterhin hoffen kann. Jonas Grethlein schreitet in dem vorliegenden Band den Horizont der Hoffnung aus und bietet zugleich einen erhellenden Durchgang durch die letzten drei Jahrtausende der westlichen Geistes- und Philosophiegeschichte. Aus allen Epochen von der Antike bis heute lässt er Philosophen und Religionsstifter, Märtyrerinnen und Literatinnen, Künstler, Opfer und Leidende zu Wort kommen. Er führt eindrücklich vor Augen, wie Menschen zu allen Zeiten darum rangen, eine ihnen unverfügbare Zukunft zu bewältigen, und zeigt neben den mannigfaltigen Formen des Hoffens auch seine Bedeutung in der Gegenwart auf. So erhellt er das Wesen der Hoffnung als Weltverhältnis, beschreibt, wie sie entstehen, worauf sie sich richten und worin sie gründen kann. „Denn Hoffnung ist das, was die meisten Menschen nährt.“ SophoklesVom Prinzip Hoffnung – ein erhellender Durchgang durch die letzten 2500 JahreBrillant dargestellt von Jonas GrethleinWarum und worauf wir trotz allem hoffen
• LegalWriter: An Intelligent Writing Support System for Structured and Persuasive Legal Case Writing for Novice Law Students. Weber, Florian; Wambsganss, Thiemo; Neshaei, Seyed Parsa; Söllner, Matthias (2024). 1052:1–1052:23.

  Novice students in law courses or students who encounter legal education face the challenge of acquiring specialized and highly concept-oriented knowledge. Structured and persuasive writing combined with the necessary domain knowledge is challenging for many learners. Recent advances in machine learning (ML) have shown the potential to support learners in complex writing tasks. To test the effects of ML-based support on students’ legal writing skills, we developed the intelligent writing support system LegalWriter. We evaluated the system’s effectiveness with 62 students. We showed that students who received intelligent writing support based on their errors wrote more structured and persuasive case solutions with a better quality of legal writing than the current benchmark. At the same time, our results demonstrated the positive effects on the students’ writing processes.
• Nachhaltige Intelligenz–intelligente Nachhaltigkeit: Transdisziplinäre Forschungsperspektiven. Hornung, G.; Blättel-Mink, B.; Helmke, J. T.; Döben-Henisch, G.-D.; Lamla, J.; Limberger, V.; Ohde, F.; Reß, B.; Schrader, C.; Söllner, M (2024).

  Nachhaltigkeit und Künstliche Intelligenz sind zwei der Megathemen unserer Zeit. Sie werfen schon je für sich komplexe Herausforderungen wissenschaftlicher und praktischer Art auf, die schon jetzt stark und immer weiter zunehmend Politik, Wirtschaft und Alltagsleben erfassen. Dem vorliegenden Whitepaper liegt die Überzeugung der Autor*innen zugrunde, dass wir trotz der jeweiligen Komplexität Nachhaltigkeit und Künstliche Intelligenz in ihren Wechselwirkungen beobachten und analysieren müssen, weil eine Trennung angesichts der vielfältigen Verflechtungen künstlich wäre und Gefahr liefe, wesentliche Problemlagen, aber auch Handlungsoptionen zu übersehen.
• KwicKwocKwac : a tool for rapidly generating concordances and marking up a literary text. Barzaghi, Sebastian; Paolucci, Francesco; Tomasi, Francesca; Vitali, Fabio (2024, October).

  This paper introduces KwicKwocKwac 1.0 (KwicKK), a web application designed to enhance the annotation and enrichment of digital texts in the humanities. KwicKK provides a user-friendly interface that enables scholars and researchers to perform semi-automatic markup of textual documents, facilitating the identification of relevant entities such as people, organizations, and locations. Key functionalities include the visualization of annotated texts using KeyWord in Context (KWIC), KeyWord Out Of Context (KWOC), and KeyWord After Context (KWAC) methodologies, alongside automatic disambiguation of generic references and integration with Wikidata for Linked Open Data connections. The application supports metadata input and offers multiple download formats, promoting accessibility and ease of use. Developed primarily for the National Edition of Aldo Moro's works, KwicKK aims to lower the technical barriers for users while fostering deeper engagement with digital scholarly resources. The architecture leverages contemporary web technologies, ensuring scalability and reliability. Future developments will explore user experience enhancements, collaborative features, and integration of additional data sources.
• Designing a Large Language Model-Based Coaching Intervention for Lifestyle Behavior Change. Meywirth, Sophia M. Mandviwalla, M. Söllner, T. Tuunanen (eds.) (2024).

  Adopting and maintaining healthy lifestyle behaviors such as regular exercise and balanced nutrition remain challenging despite their well-documented benefits for preventing chronic diseases and promoting overall well-being. Motivational Interviewing (MI) has emerged as a promising technique to address ambivalence and facilitate behavior change. However, traditional face-to-face delivery of MI interventions is limited by scalability and accessibility issues. Leveraging recent advancements in LLMs, this paper proposes an innovative approach to deliver MI-based coaching for lifestyle behavior change digitally. Following a problem-centered DSR approach, we created an initial prototype based on MI theory and qualitative user interviews using ChatGPT (GPT-3.5). We evaluated our prototype in a qualitative study. Our research outcomes include five design principles and thirteen system requirements. This research enhances the design knowledge base in LLM-based health coaching. It marks an essential first step towards designing LLM-based MI interventions, contributing valuable insights for future research in this emerging field.
• Behavior Change Techniques in Digital Interventions for Health Behavior Change: A Taxonomy of Popular Mobile Applications. Meywirth, Sophia; Söllner, Matthias (2024).

  Health behavior change interventions can support individuals in establishing habits beneficial to their health. Hence, they can help to prevent diseases and reduce mortality caused by unhealthy behaviors. The delivery of such interventions through mobile phones is promising. We developed a taxonomy of popular mobile applications for health behavior change, identifying the intervention’s aim and applied behavior change techniques. We classified 63 popular mobile applications according to our taxonomy. Thereby identifying the most prevalent behavior change techniques and combinations of the application’s aim and applied behavior change techniques. This result can be used to inform the design of future applications.
• Use IT again? Dynamic roles of habit, intention and their interaction on continued system use by individuals in utilitarian, volitional contexts. Söllner, Matthias; Mishra, Abhay Nath; Becker, Jan-Michael; Leimeister, Jan Marco (2024). 33(1) 80–96.

  This paper employs a longitudinal perspective to examine continued system use (CSU) by individuals in utilitarian, volitional contexts when alternative systems are present . We focus on two key behavioural antecedents of CSU – habit and continuance intention – and theorise how the relationships between CSU and these antecedents evolve over time. In addition, we hypothesise how the interaction effect of habit and intention on CSU evolves temporally. Our theorising differs from extant literature in two important respects: 1) In contrast to the widespread acceptance of the diminishing effect of continuance intention on CSU in the information systems (IS) literature, we hypothesise that in our context, its impact increases with time; and 2) In contrast to the negative moderation effect of habit on the relationship between intention and CSU proposed in the literature, we posit a positive interaction effect. We collect longitudinal survey data on the use of a higher education IS from students in a European university. Our results suggest that the impact of continuance intention on CSU as well as the interaction effect between habit and intention are increasing over time. We further introduce a methodological innovation – the permutation approach to conduct the multi-group analysis with repeated measures – to the literature.
• Navigating the Writing Mind: Intelligent Feedback to Supporting Students’ Cognitive Writing Process. Weber, F. (2024). 1–16.

  Advances in natural language processing (NLP) and machine learning (ML) have resulted in numerous designs for learning systems in the IS community, which offer learners intelligent feedback during the writing process. Despite the increasing interest in research on learning systems for writing skills, there still needs to be more empirical knowledge regarding the impacts of intelligent feedback in different phases of cognitive writing. Therefore, my objective is to explore the effects of intelligent feedback on students' writing skills. I center my investigation on the cognitive writing model, examining how intelligent feedback influences student outcomes. Through a field experiment involving 101 students, I demonstrate the positive effects of intelligent feedback on students' writing skills across all stages of writing (planning, translating (writing), and revision). In addition, I was able to gain initial insights into the phases of writing in which intelligent feedback is most effective.
• Modelling Argumentation for an User Opinion Aggregation Tool. Weingart, Pablo; Wambsganss, Thiemo; Söllner, Matthias (2024). 11548–11559.

  We introduce an argumentation annotation scheme that models basic argumentative structure and additional contextual details across diverse user opinion domains. Drawing from established argumentation modeling approaches and related theory on user opinions, the scheme integrates the concepts of argumentative components, specificity, sentiment and aspects of the user opinion domain. Our freely available dataset includes 1,016 user opinions with 7,266 sentences, spanning products from 19 e-commerce categories, restaurants, hotels, local services, and mobile applications. Utilizing the dataset, we trained three transformer-based models, demonstrating their efficacy in predicting the annotated classes for identifying argumentative statements and contextual details from user opinion documents. Finally, we evaluate a prototypical dashboard that integrates the model inferences to aggregate information and rank exemplary products based on a vast array of user opinions. Early results from an experimental evaluation with eighteen users include positive user perceptions but also highlight challenges when condensing detailed argumentative information to users.
• Towards Effective AI-Driven Reading Assistants: A Design Science Exploration. Thirty-Second European Conference on Information Systems. Hänel, Martin; Wambsganss, Thiemo; Söllner, Matthias (2024).

  Recent advancements in AI have led to the introduction of tools that support researchers in scientific reading. Tools such as SciSpace have come to the forefront to assist users in reading scientific texts. However, there is an insufficient theoretical foundation on how to design these reading assistants as well as no evidence of their effects, especially given the recent progress. Specifically, past literature lacks insights on evaluated user requirements and design principles for the design of computer-assisted reading systems. Addressing these challenges, we draw on Design Science Research (DSR) to derive and evaluate a set of five design principles for computer-assisted reading systems. Building on flow theory as our theoretical lens, we develop and perform a first proof-of-concept evaluation of a prototypical implementation of our principles as a computer-assisted reading artifact. Our design principles support researchers and practitioners on how to design, evaluate, and compare their AI-reading tools more effectively.
• Enlaces Abril. me (2024).
• Gibraltar subduction zone is invading the Atlantic. Duarte, João C.; Riel, Nicolas; Rosas, Filipe M.; Popov, Anton; Schuler, Christian; Kaus, Boris J.P (2024). 52(5) 331–335.

  Subduction initiation is a cornerstone of the Wilson cycle. It marks the turning point in an ocean's lifetime, allowing its lithosphere to be recycled into the mantle. However, formation of new subduction zones in Atlantic-type oceans is challenging, given that it commonly involves the action of an external force, such as the slab pull from a nearby subduction zone, a far-field compression, or the impact of a plume. Notwithstanding, the Atlantic already has two subduction zones, the Lesser Antilles and the Scotia arcs. These subduction zones have been forced from the nearby Pacific subduction zones. The Gibraltar arc is another place where a subduction zone is invading the Atlantic. This corresponds to a direct migration of a subduction zone that developed in the closing Mediterranean Basin. Nevertheless, few authors consider the Gibraltar subduction to be still active because it has significantly slowed down in the past millions of years. Here, we use new gravity-driven geodynamic models that reproduce the evolution of the Western Mediterranean, show how the Gibraltar arc formed, and test if it is still active. The results suggest that the arc will propagate further into the Atlantic after a period of quiescence. The models also show how a subduction zone starting in a closing ocean (Ligurian Ocean) can migrate into a new opening ocean (Atlantic) through a narrow oceanic corridor. Subduction invasion is likely a common mechanism of subduction initiation in Atlantic-type oceans and a fundamental process in the recent geological evolution of Earth.
• Theory-Driven Design of a Negotiation Canvas for Reaching Win-Win Agreements. Zahn, Eva-Maria; Dickhaut, Ernestine in Lecture Notes in Computer Science (2024). (Vol. 14621) 123–138.

  In today’s interconnected world, negotiations are essential for navigating the complexities of personal and professional life. Resulting in a “negotiation network”, integrative negotiation strategies are gaining relevance for mutual benefit. However, negotiators often fall short due to inadequate preparation, focusing more on their own negotiation goals than understanding the other party’s priorities and interests. This challenge stems from inadequate preparation which is crucial in the overall negotiation process. A comprehensive preparation empowers negotiators to conduct a stronger and more reflective negotiation and to enable unexpected challenges with more confidence. As a result, negotiators can make informed and strategic decisions. However, an artifact that explicitly addresses the preparation process and prepares inexperienced negotiators in a structured and theoretically founded manner is missing. Therefore, we introduce a theory-driven negotiation canvas to prepare negotiations and reach win-win agreements. The canvas, addressing crucial negotiation elements, offers a structured approach to preparing for an upcoming negotiation. This research contributes to a more comprehensive understanding of negotiation preparation and provides a valuable tool to empower negotiators.
• Let’s Chat to Negotiate: Designing a Conversational Agent for Learning Negotiation Skills. Zahn, E.-M.; Schöbel, S. M. Mandviwalla, M. Söllner, T. Tuunanen (eds.) (2024). (Vol. 14621) 229–243.
• Mapping soft skills and further research directions for higher education: A bibliometric approach with structural topic modelling. Zahn, E.-M.; Schöbel, S.; Saqr, M.; Söllner, M. (2024). 1–21.
• Technology-based Learning Services for Transferable Skills – A Systematic Literature Review. Zahn, E.-M.; Schöbel, S.; Söllner, M. (2024).
• Hurricane‐induced pollinator shifts in a tightly coadapted plant–hummingbird mutualism. Schrøder, Taia S. O.; Gonçalves, Fernando; Vollstädt, Maximilian G. R.; Zhang, Tianying; Jensen, Rasmus D.; Tarazona‐Tubens, Fabio L.; Kim, Seokmin; Galetti, Mauro; Simmons, Benno I.; Kaiser‐Bunbury, Christopher N.; Temeles, Ethan J.; Dalsgaard, Bo (2024). 244(1) 16–20.
• Employee acceptance of digital transformation strategies: A paradox perspective. Klein, Sascha P.; Spieth, Patrick; Söllner, Matthias (2024). 41(5) 999–1021.

  Abstract Digital transformation (DT) strategies often aim at innovating an organization's business models (BMs) and developing digital innovations. However, most of the DT strategies fail or result only in incremental innovation. Research predominantly identified critical management capabilities for DT success, neglecting the role of employees, although employee support is vital for the successful implementation of DT strategies. We conceptualize employee acceptance of DT strategies and draw on paradox theory and organizational change literature to shed light on the antecedents of employee acceptance. Using survey-based data from an incumbent introducing a DT strategy, we find empirical support that employees with a paradox mindset are likely to accept the DT strategy regardless of the expected scope of change, while a high scope of expected changes weakens the effect of a positive attitude toward change on acceptance. The findings contribute to understanding the micro-level aspects of DT and extend research emphasizing top-down management approaches in DT with an employee perspective. This study extends previous findings in innovation management that predominantly provide evidence on the acceptance of specific digital technologies and offer insights into the antecedents of DT strategy acceptance. We offer managers insights into how employees perceive DT strategies, which can help to leverage the potential of digital innovation.
• Google, Microsoft, and Perplexity are promoting scientific racism in search results. Gilbert, David (2024).

  The web’s biggest AI-powered search engines are featuring the widely debunked idea that white people are genetically superior to other races.
• Neustart in Zeiten des Fachkräftemangels : die Fusion zur RPTU Kaiserslautern-Landau. Chancen und Herausforderungen. Wollschläger, Thomas; Heisel, Carina (2024). 11(4) 1–6.
• Synthetic human memories : AI-edited images and videos can implant false memories and distort recollection. Pataranutaporn, Pat; Archiwaranguprok, Chayapatr; Chan, Samantha W. T.; Loftus, Elizabeth; Maes, Pattie (2024, September).

  AI is increasingly used to enhance images and videos, both intentionally and unintentionally. As AI editing tools become more integrated into smartphones, users can modify or animate photos into realistic videos. This study examines the impact of AI-altered visuals on false memories--recollections of events that didn't occur or deviate from reality. In a pre-registered study, 200 participants were divided into four conditions of 50 each. Participants viewed original images, completed a filler task, then saw stimuli corresponding to their assigned condition: unedited images, AI-edited images, AI-generated videos, or AI-generated videos of AI-edited images. AI-edited visuals significantly increased false recollections, with AI-generated videos of AI-edited images having the strongest effect (2.05x compared to control). Confidence in false memories was also highest for this condition (1.19x compared to control). We discuss potential applications in HCI, such as therapeutic memory reframing, and challenges in ethical, legal, political, and societal domains.
• The geometric structure of topic models. Hirth, Johannes; Hanika, Tom (2024, March).

  Topic models are a popular tool for clustering and analyzing textual data. They allow texts to be classified on the basis of their affiliation to the previously calculated topics. Despite their widespread use in research and application, an in-depth analysis of topic models is still an open research topic. State-of-the-art methods for interpreting topic models are based on simple visualizations, such as similarity matrices, top-term lists or embeddings, which are limited to a maximum of three dimensions. In this paper, we propose an incidence-geometric method for deriving an ordinal structure from flat topic models, such as non-negative matrix factorization. These enable the analysis of the topic model in a higher (order) dimension and the possibility of extracting conceptual relationships between several topics at once. Due to the use of conceptual scaling, our approach does not introduce any artificial topical relationships, such as artifacts of feature compression. Based on our findings, we present a new visualization paradigm for concept hierarchies based on ordinal motifs. These allow for a top-down view on topic spaces. We introduce and demonstrate the applicability of our approach based on a topic model derived from a corpus of scientific papers taken from 32 top machine learning venues.
• NEXUS : eine kurze Geschichte der Informationsnetzwerke von der Steinzeit bis zur künstlichen Intelligenz Harari, Yuval Noah (2024). Penguin Verlag, München.

  In den letzten 100 000 Jahren hat die Menschheit enorme Macht erlangt. Doch trotz all unserer Entdeckungen, Erfindungen und Eroberungen befinden wir uns heute in einer existenziellen Krise. Die Welt steht am Rande des ökologischen Zusammenbruchs. Zuhauf werden Falschinformationen verbreitet. Und wir stürzen uns kopfüber in das Zeitalter der künstlichen Intelligenz – ein neues Informationsnetzwerk, das uns auszulöschen droht. Wenn wir so klug sind, warum sind wir dann so selbstzerstörerisch? »Nexus« zeigt, wie der Informationsfluss uns und unsere Welt geformt hat. Yuval Noah Harari nimmt uns mit von der Steinzeit und biblischen Zeiten über die frühneuzeitlichen Hexenverfolgungen, den Stalinismus und den Nationalsozialismus bis zum Wiederaufleben des Populismus in der heutigen Zeit. Dabei lenkt er unseren Blick auf die komplexe Beziehung zwischen Information und Wahrheit, Bürokratie und Mythologie, Weisheit und Macht. Er erkundet, wie verschiedene Gesellschaften und politische Systeme Informationen genutzt haben, um ihre Ziele zu erreichen – zum Guten wie zum Schlechten. Und er befasst sich mit den drängenden Entscheidungen, vor denen wir heute stehen, da nicht-menschliche Intelligenz unsere Existenz bedroht. Informationen sind nicht der Rohstoff, aus dem die Wahrheit ist, aber auch nicht einfach nur eine Waffe. »Nexus« erkundet den hoffnungsvollen Mittelweg zwischen diesen Extremen und zeigt, wie sich unser gemeinsames Menschsein wiederentdecken lässt.
• AI : a means to an end or a means to our end?. Fry, Stephen (2024, September).

  The text of a talk I gave on Thursday 12th September as the inaugural “Living Well With Technology” lecture for King’s College London’s Digital Futures Institute.
• Atlas der KI : die materielle Wahrheit hinter den neuen Datenimperien Crawford, Kate (2024). C.H. Beck, München.

  Wir tendieren dazu, künstliche Intelligenz als eine wundersame und körperlose Form der maschinellen Klugheit zu betrachten. Von der preisgekrönten Wissenschaftlerin Kate Crawford lernen wir hingegen, dass KI in Wahrheit weder künstlich noch intelligent ist, sondern in ihrer materiellen Wirklichkeit auf Ressourcenausbeutung und Machtkonzentration hinausläuft. Crawford nimmt uns mit auf eine faszinierende Reise zu Lithiumminen und Klickfabriken, zu automatisierten Arbeitsplätzen und riesigen Datenarchiven, zu AI-Trainingscamps und zum algorithmischen Kriegsführungsteam des Pentagon. Auf diese Weise zeichnet sie einen Atlas der künstlichen Intelligenz, der die verschiedenen Bereiche ihrer konkreten Realität kartiert, um unser kritisches Auge zu schulen. Gestützt auf ein Jahrzehnt originärer Forschung zeigt Crawford, dass KI in erster Linie eine Technologie der Extraktion ist – der Abschöpfung von Mineralien, billiger Arbeitskraft und einer unermesslichen Anzahl von Daten. Das planetare Netzwerk der KI schädigt unsere Umwelt massiv, vertieft soziale Ungleichheiten und bedroht demokratische Prinzipien. Crawfords Buch liefert uns einen dringlichen Bericht, was auf dem Spiel steht, wenn große Unternehmen und staatliche Institutionen KI nutzen, um die Welt umzugestalten. „Ein wertvolles Korrektiv für den Hype um KI und eine nützliche Gebrauchsanweisung für die Zukunft.“ John Thornhill, Financial Times Ein neuer Blick auf das, was künstliche Intelligenz in Wirklichkeit ist Aus welchen Materialen bestehen KI-Systeme, und wo kommen die her? Welche Formen menschlicher Arbeit, welche Wissenssysteme und Machtverhältnisse müssen dabei am Start sein? Crawford zeigt: KI ist die ausbeuterische Industrie des 21. Jahrhunderts KI hat in ökologischer, sozialer und politischer Hinsicht verheerende Wirkungen
• Bellizismus und Pazifismus : Moralisieren vs. Moral. Ruschig, Ulrich (2024). 12–13.

  Rechtfertigungen heutiger Kriegsvorbereitung und die Aktualität von Immanuel Kants »Zum ewigen Frieden«
• Personal identity and the Self. Madden, Rory (2024).

  Cambridge Core - Philosophy of Mind and Language - Personal Identity and the Self
• Discovery und Geodaten : neue explorative Wege mit Graphen und Vektoren. Uttenweiler, Bernd (2024). 11(4) 1–9.
• A review of state of the art deep learning models for ontology construction. Zengeya, Tsitsi; Vincent Fonou-Dombeu, Jean (2024). 12 82354–82383.

  Researchers are working towards automation of ontology construction to manage the ever-growing data on the web. Currently, there is a shift from the use of machine learning techniques towards exploration of deep learning models for ontology construction. Deep learning model are capable of extracting terms, entities, relations, and classifications, and perform axiom learning from the underutilized richness of web-based knowledge. There has been remarkable progress in automatic ontology creation using deep learning models since they can perform word embedding, long-term dependency acquisition, concept extraction from large corpora, and inference of abstracted relationships based on broad corpora. Despite their emerging importance, deep learning models remain underutilized in ontology construction, and there is no comprehensive review of their application in ontology learning. This paper presents a comprehensive review of existing deep-learning models for the construction of ontologies, the strength and the weaknesses presented by the deep learning models for ontology learning as well as promising directions to achieve a more robust deep learning models. The Deep Learning models reviewed include Recurrent Neural Networks (RNNs), Convolutional Neural Networks (CNNs), Long-Short Term Memory (LSTMs), and Gated Recurrent Unit (GRU) as well as their ensembles. While these traditional deep learning models have achieved great success, one of their limitations is that they struggle to understand the meaning and order of data in sequences. CNNs and RNN-based models such as LSTMs and GRUs can be computationally expensive due to their large number of parameters or complex gating mechanisms. Furthermore, RNN models suffer from vanishing gradients, making it difficult to learn long-term relationships in sequences. Additionally, RNN-based models process information sequentially, limiting their ability to take advantage of powerful parallel computing hardware, slowing down training and inference, especially for long sequences. Consequently, there has been a shift towards Generative Pre-Trained (GPT) models and Bidirectional Encoder Representations from Transformers (BERT) models. This paper also reviewed the GPT-3, GPT-4, and the BERT models for extracting terms, entities, relations, and classifications. While GPT models excel in contextual understanding and flexibility, they fall short when handling domain-specific terminology and disambiguating complex relationships. Fine-tuning and domain-specific training data could minimize these shortcomings, and further enhance the performance of GPT in term and relation extraction tasks. On the other hand, the BERT models excel in comprehending context-heavy texts, but struggles with higher-level abstraction and inference tasks due to a lack of explicit semantic knowledge, thus necessitating inference for unspecified relationships. The paper recommends further research on deep learning models for ontology alignment and merging. Also, the ensembling of deep learning models and the use of domain-specific knowledge for ontology learning require further research for ontology construction.
• Zensurvorwürfe in der Genderforschung. Frei, Martina (2024, October).

  Eine Studie zu plötzlicher Geschlechtsidentitätsstörung löst einen Sturm der Entrüstung aus – und wird vom Verlag zurückgezogen.
• Leveraging Low Code Development of Smart Personal Assistants: An Integrated Design Approach with the SPADE Method. Elshan, Edona; Ebel, Philipp Alexander; Söllner, Matthias; Leimeister, Jan Marco (2023). 40(1) 96–129.

  Smart personal assistants (SPAs), such as Alexa for example, promise individualized user interactions owing to their varying interaction possibilities, knowledgeability, and human-like behaviors. To support the widespread adoption and use of SPAs, organizations such as Google or Amazon provide low code environments that support the development of SPAs (e.g., for Google Home or Amazon’s Alexa). These so-called low code platforms enable domain experts (e.g., business users without programming skills or experience) to develop SPAs for their purposes. However, using these platforms alone does not guarantee a useful and good conversation with novel SPAs due to non-intuitive design choices. Following a design science research approach, we propose the Smart Personal Assistant for Domain Experts (SPADE) method to address the missing link. This method supports domain experts in the development and contextualization of sophisticated SPAs for various application scenarios and focuses especially on conversational and anthropomorphic design steps. Our proof of concept and proof of value results show that SPADE is useful for supporting domain experts to create effective SPAs in different domains beyond private set-ups.
• ChatGPT und andere »Quatschmaschinen« : Gespräche mit Künstlicher Intelligenz Tuschling, Anna; Sudmann, Andreas; Dotzler, Bernhard J. (2023). transcript Verlag, Bielefeld.

  Die Veröffentlichung von ChatGPT im Herbst 2022 heizte die Kontroverse um Künstliche Intelligenz an und führte zu einer seitdem unaufhörlichen Fragelust - verstärkt dadurch, dass dieselben Prompts schon kürzeste Zeit später andere Outputs generieren. In einem experimentellen Format präsentieren die Herausgeber*innen erste kommentierte Gespräche mit KI-Sprachmodellen. Sie geben Einblick in dialogische Szenen, die eine fortlaufende Transformation der Technik und Eigentümlichkeiten maschinellen Lernens erfassen. Die Sammlung zielt in Form witziger, unheimlicher und - mehr oder weniger - kluger Dialoge zwischen Mensch und Maschine auf die Dokumentation einer mediengeschichtlichen Passage in eine neue Ära allgegenwärtiger KI.
• Vom Wald : eine Philosophie der Freiheit Grau, Alexander (2023). (1st ed.) Claudius, München.

  Wald ist Freiheit. Wald schafft Distanz zum Alltag. Er schirmt ab und lenkt den Blick auf unser Menschsein. Er ist der Ort, an dem man sich von den Manipulationen und Einflüsterungen des Zeitgeistes lösen kann. Er gibt uns Raum für unser Bewusstsein und lässt uns zu uns kommen. Und er konfrontiert uns mit dem Ursprung menschlicher Existenz. In seinem eindringlichen Essay entwirft Alexander Grau eine zeitgemäße Philosophie des Waldes jenseits von Naturkitsch, Wellness und Esoterik. Dabei führt er durch die Kultur- und Ökologiegeschichte des Waldes ebenso wie durch die literarischen und philosophischen Projektionen, die unser Bild vom Wald in der Moderne geprägt haben.
• Ich, das Gehirn : zum Paradox eines sich selbst erforschenden Systems Hurka, Herbert M. (2023). transcript Verlag, Bielefeld.

  Vieles deutet darauf hin, dass das 21. Jahrhundert die Epoche des Gehirns werden könnte - samt einer wissenschaftlichen und kulturellen Revolution. Die anstehenden Veränderungen im Zusammenhang mit Künstlicher Intelligenz und der aufstrebenden Weltanschauung des Transhumanismus werfen die grundsätzliche Frage auf, ob ein System in der Lage ist, sich selbst zu durchschauen. In einer essayistischen Aufschlüsselung dieser Problematik widmet sich Herbert M. Hurka der Selbstthematisierung des Gehirns und zeigt: Mit jeder Äußerung, gleich in welchem Medium, betreibt das Gehirn in einer selbstreferentiellen Schleife immer auch aktiv seine Selbsterforschung - und bleibt dabei als Subjekt der Erkenntnis immer auch ihr Objekt.
• Die Macht der Computermodelle : Quellen der Erkenntnis oder digitale Orakel? Simeon, Bernd (2023). (1. Aufl. 2023 Editionst ed.) Springer, Berlin, Heidelberg.

  Unsichtbar und zugleich omnipräsent, erobern sich Computermodelle ständig neue Anwendungsfelder. Ihre Macht verdanken sie den immensen Kapazitäten moderner Rechnersysteme, und sie wird zusätzlich befeuert durch das maschinelle Lernen aus riesigen Datenbergen. Digitale Patienten, Szenarien zum Klimawandel, Missionen zu fernen Planeten, das Jonglieren an den Finanzmärkten, die Entwicklung nuklearer Waffensysteme – anhand vielfältiger Episoden und persönlicher Erlebnisse des Autors bietet Ihnen dieses Buch Einblick in eine faszinierende Welt, die zu einer wesentlichen Quelle des Erkenntnisgewinns geworden ist. Doch die unaufhaltsame Mathematisierung schürt auch Ängste: Sind wir, wie Goethes Zauberlehrling, schon tief im Netz der Algorithmen und Computermodelle verstrickt und verlieren die Kontrolle über ihre Macht? Wie ein modernes Orakel stillen Computermodelle unsere Sehnsucht nach einer berechenbaren Zukunft. Doch es lauern Fallstricke, aufgrund derer ihre Prognosen gravierend danebenliegen können. Blindes Vertrauen beim Laien wie auch starre Modellgläubigkeit beim Experten resultieren dann in Fehlentscheidungen, dem Versagen technischer Systeme oder ökonomischem Desaster. Kann man diesen Fallstricken entgehen?Zum Lesen dieses Buches benötigen Sie Neugierde sowie Offenheit für Überraschendes, aber keine Fachkenntnisse. Eine Fülle von Abbildungen veranschaulicht die wesentlichen Zusammenhänge, ergänzt um spannende Anekdoten aus Wissenschaft und Technik. Wollen Sie mehr erfahren? Dann kommen Sie mit auf diesen Streifzug durch den Kosmos der Computermodelle.
• Nudging Digital Learning – An Experimental Analysis of Social Nudges to Manage Self-Regulated Learning and Online Learning Success. Schlegel, Laura; Schöbel, Sofia; Söllner, Matthias (2023).

  Self-regulated learning competencies are of increasing importance to ensure learning success in online learning environments. We investigate the use of digital social nudges in a self-reliant online learning situation to support learners in better managing their self-regulated learning behaviors. We ground our research on dual-process theory and social comparison theory to design social nudges. To evaluate our research model, we conduct an online experiment (N=226). The results show that social nudges positively impact learning outcomes mediated by self-regulated learning behaviors manifested using learning strategies. We found that positive emotions can further strengthen the positive effect of social nudges. Our results help to understand how social nudges can be efficiently used in online learning environments to support learners in better managing their learning processes and achieving learning outcomes. We open new chances for researchers and designers of online learning materials to support online learning processes.
• Hybris und Nemesis : wie uns die Entzivilisierung von Macht in den Abgrund führt ; Einsichten aus 5000 Jahren Mausfeld, Rainer (2023). (1st ed.) Westend, Neu-Isenburg.

  Macht drängt nach mehr Macht und Reichtum nach mehr Reichtum, eine Dynamik, die den Zusammenhalt einer Gesellschaft gefährdet und sie zu zerstören droht: Dies ist eine der frühesten Einsichten der Zivilisationsgeschichte. Macht bedarf daher stets einer robusten Einhegung. Das bedeutendste Schutzinstrument für eine Zivilisierung von Macht stellt die egalitäre Leitidee der Demokratie dar. Rainer Mausfeld zeigt entlang historischer Linien auf, dass der Begriff der Demokratie seiner ursprünglichen Bedeutung beraubt worden ist und heute als Demokratierhetorik für Herrschaftszwecke missbraucht wird. Dadurch ist es in den vergangenen Jahrzehnten zu einer Entzivilisierung von Macht gekommen, deren psychische, gesellschaftliche und ökologische Auswirkungen die menschliche Zivilisation insgesamt bedrohen.
• A speech-based empathy training system - initial design insights. Weber, Florian; Schween, Sebastian; Wambsganß, Thiemo; Söllner, Matthias (2022).

  Empathy is an essential component of human communication since it increases our understanding and perception of others. However, studies show that students' empathy skills have declined rapidly in the last decades. Against this background, practitioner reports predict that the importance of empathy will increase as a skill for successful agile teamwork in the future. Therefore, researchers have designed information systems to train empathy abilities of learners in different domains. Nevertheless, research on automated speech-based training is rather scarce. Hence, we aim to investigate how to design a speech-based empathy training system that helps students react emotionally adequately in communication. This research in progress paper presents five initial requirements that guide future research and development of a speech-based empathy training system intended to support students' self-regulated learning. With this, we hope to provide guidance for the design and embedding of speech-based empathy training systems at scale.
• Kollaboration im Datenschutz. Datenschutz und Datensicherheit. Hornung, Gerrit; Söllner, Matthias; Stroscher, Jan-Philipp; Zahn, Eva-Maria (2022). 46(6) 384–389.

  Die DSGVO enthält sehr unterschiedliche Vorgaben zur Zusammenarbeit verschiedener Akteure, aber kein systematisches Modell für die Ausgestaltung dieser Formen der Zusammenarbeit. Gerade für KMU könnten standardisierte Verfahren ein Ansatz zur effizienten Umsetzung sein. Der Beitrag untersucht die Frage, an welchen Stellen Ansätze des Collaboration Engineering Zusammenarbeitsprozesse ermöglichen würden, die die gesetzlichen Vorgaben mit Leben füllen.
• Estimating dispersal using close kin dyads: The kindisperse R package. Jasper, Moshe E.; Hoffmann, Ary A.; Schmidt, Thomas L. (2022). 22(3) 1200–1212.

  Abstract Investigating dispersal in animal populations can be difficult, particularly for taxa that are hard to directly observe such as those that are small or rare. A promising solution may come from new approaches that use genome-wide sequence data to detect close kin dyads and estimate dispersal parameters from the distribution of these dyads. These methods have so far only been applied to mosquito populations. However, they should have broad applicability to a range of taxa, although no assessment has yet been made on their performance under different dispersal conditions and study designs. Here we develop an R package and shiny app, kindisperse, that can be used to estimate dispersal parameters from the spatial distribution of close kin. kindisperse can handle study designs that target different life stages and allows for a range of dispersal kernel shapes and organismal life histories; we provide implementation examples for a vertebrate (Antechinus) and an invertebrate (Aedes). We use simulations run in kindisperse to compare the performance of two published close kin methodologies, showing that one method produces unbiased estimates whereas the other produces downward-biased estimates. We also use kindisperse simulations to investigate how study design affects dispersal estimates, and we provide guidelines for the size and shape of sample sites as well as the number of close kin needed for accurate estimates. kindisperse is easily adaptable for application to a variety of research contexts ranging from invasive pests to threatened species where noninvasive DNA sampling can be used to detect close kin.
• Feminismus, Algorithmen, Gender-Data-Gap und was das alles mit Bibliotheks- und Informationswissenschaft zu tun hat. Juen, Sara (2021). (39)

  In den letzten Jahren gab es vermehrt Berichte über unethisches Handeln in Technologiekonzernen und Hinweise darauf, dass Algorithmen nicht neutral handeln, sondern gesellschaftlich bestehende Ungerechtigkeiten weiterführen oder diese sogar verschärfen. Der vorliegende Artikel gibt einen Einblick in diese Thematik und zeigt auf, warum dieses Thema für die Bibliotheks- und Informationswissenschaft relevant ist.
• Zwischen Normierung und Offenheit : Potenziale und offene Fragen bezüglich kontrollierter Vokabulare und Normdateien. Strickert, Moritz (2021). (40)

  Der Rückgriff auf kontrollierte Vokabulare und Normdaten ist im Zuge der Zugänglichmachung von analogen und digitalen Ressourcen zentral. Der vorliegende Text diskutiert (ethisch-theoretische) Fragestellungen im Umgang mit kontrollierten Vokabularen als spezifisch eurozentrische Wissensordnungen.
• Accurate Numerical Method for Singular Initial-Value Problems. Tesfaye Aga Bullo1*, Gemechis File Duressa1; Kiltu2, Gashu Gadisa S. Dixon, D. Bainbridge, R. Typke (eds.) (2021).
• Semantik und künstliche Intelligenz : Beiträge zur automatischen Sprachbearbeitung II. Eisenberg, Peter in De Gruyter Studienbuch (2016). De Gruyter.
• What’s the frequency, Kenneth? : eine (queer)feministische Kritik an Sexismen und Rassismen im Schlagwortkatalog. Sparber, Sandra (2016). 69(2) 236–243.

  Sprachliche Diskriminierungen bilden sich auf verschiedenen Ebenen im Schlagwortkatalog ab. In unterschiedlichen Ausprägungen finden wir sie sowohl in den Normdaten als auch in der Schlagwortfolge. Ziel des Aufsatzes ist es, die bibliothekarische Inhaltserschließung als soziale Praxis in einen theoretischen Bezugsrahmen einzubetten, um sie einer fundierten Kritik zugänglich zu machen
• Das Geschlecht spukt in der Stadtbibliothek : ein Aufruf für genderneutrale Bibliotheksangebote. Leyrer, Katharina (2014). (25) 76–79.
• Die Frau im Bibliothekskatalog. Aleksander, Karin (2014). (25) 8–16.
• Das Zeitalter der Erkenntnis : die Erforschung des Unbewussten in Kunst, Geist und Gehirn von der Wiener Moderne bis heute Kandel, Eric (2012). (4th ed.) Siedler Verlag, München.

  Was passiert in unserem Gehirn, wenn wir Kunst betrachten?Nobelpreisträger Eric Kandel hat mit »Das Zeitalter der Erkenntnis« ein brillantes Buch geschrieben, das uns in das Wien Sigmund Freuds, Gustav Klimts und Arthur Schnitzlers entführt. Dort setzten um 1900 die angesehensten Köpfe der Naturwissenschaft, Medizin und Kunst eine Revolution in Gang, die den Blick auf den menschlichen Geist und seine Beziehung zur Kunst für immer verändern sollte.Was passiert in unserem Gehirn, wenn wir Kunst betrachten, welche Einsichten und Emotionen werden dadurch ausgelöst? Und inwiefern sind sie Grundlage für Empathie oder Kreativität? Um das herauszufinden, begibt sich Eric Kandel in das Wien um 1900, wo ein enger und produktiver Austausch zwischen den bedeutendsten Psychologen, Künstlern und Literaten der Zeit stattfand. Indem Kandel diese ganz besondere Atmosphäre wiederauferstehen lässt, erzählt er die Geschichte der modernen Wissenschaft des Geistes von ihren Anfängen bis heute.
• Die Nicht-Gleichbehandlung von Frauen und Männern in der Schlagwortnormdatei - ein Offener Brief. Jank, Dagmar (1991). 25(9) 1418–1421.

  Der Artikel Die Nicht-Gleichbehandlung von Frauen und Männern in der Schlagwortnormdatei - ein Offener Brief wurde am 1. September 1991 in der Zeitschrift Bibliotheksdienst (Band 25, Heft 9) veröffentlicht.