Swiss Distance University of Applied Sciences

Online learning platforms have expanded access to education but also raise concerns about biased content, particularly in text-based learning materials such as textbooks, lesson plans, and course excerpts. Such biases can perpetuate discrimination, can harm student outcomes, and can often be difficult to detect, as identification typically relies on time-consuming human review. Learning analytics (LA) can enhance this process by supporting human reviewers through automated detection, offering a scalable solution while retaining human judgment for nuanced evaluations. Accordingly, this LA study explores two research questions: RQ1: Which features might support the identification of ethnic bias in text-based online learning materials? and RQ2: Which classification approaches might be suitable for identifying ethnic bias in text-based online learning materials? First, we identified features signalling potential ethnic bias (presence or absence) in textual content using a dataset (N = 345) labelled by 193 students from diverse ethnic backgrounds. Then, we evaluated multiple machine learning (ML) models for their effectiveness in bias classification. The results suggest significant correlations between perceived bias and content from social sciences. Additionally, through bootstrap analysis, support vector machines and random forest classifiers showed consistent performance in bias identification (with F1-scores of 0.71 and 0.70 on the test set, respectively). In contrast, the naive Bayes (NB) model demonstrated the highest precision (0.75 on the test set). We discuss these findings and their implications for LA, emphasizing the importance of quality and inclusive educational tools. As an initial step toward automated bias classification in education, this study provides a foundation for spotting ethnic bias in learning content, supporting fairer technologies for more inclusive learning environments.

Let Pi be a regular algebraic cuspidal automorphic representation (RACAR) of textrm{GL}_3(mathbb {A}_{mathbb {Q}}). When Pi is p-nearly-ordinary for the maximal standard parabolic with Levi textrm{GL}_1 times textrm{GL}_2, we construct a p-adic L-function for Pi . More precisely, we construct a (single) bounded measure L_p(Pi ) on mathbb {Z}_p^{times } attached to Pi , and show it interpolates all the critical values L(Pi times eta ,-j) at p in the left-half of the critical strip for Pi (for varying eta and j). This proves conjectures of Coates–Perrin-Riou and Panchishkin in this case. We also prove a corresponding result in the right half of the critical strip, assuming near-ordinarity for the other maximal standard parabolic. Our construction uses the theory of spherical varieties to build a “Betti Euler system”, a norm-compatible system of classes in the Betti cohomology of a locally symmetric space for textrm{GL}_3. We work in arbitrary cohomological weight, allow arbitrary ramification at p along the Levi factor of the standard parabolic, and make no self-duality assumption. We thus give the first constructions of p-adic L-functions for RACARs of textrm{GL}_n(mathbb {A}_{mathbb {Q}}) of ‘general type’ (i.e. those that do not arise as functorial lifts) for any n >2.

ABSTRACTIn this note, we analyze the performance of a simple root‐finding algorithm in uniform attachment trees. The leaf‐stripping algorithm recursively removes all leaves of the tree for a carefully chosen number of rounds. We show that, with probability , the set of remaining vertices contains the root and has a size only depending on but not on the size of the tree.

Abstract There is growing evidence supporting the usefulness of immersive virtual reality (IVR) for collaboration. However, its widespread adoption in the workplace is still pending. For a successful introduction to the professional environment, it is crucial to identify the factors influencing the acceptance of IVR for collaboration. Therefore, in this study, we employed an adapted UTAUT2 technology acceptance model to examine the influence of performance expectancy, effort expectancy, facilitating conditions, social influence, and hedonic motivation on the behavioral intention to use IVR. Conducting and analyzing an online survey of 273 participants using structural equation modeling showed that the employed technology acceptance model provided a strong fit to the survey data. Furthermore, we found that performance expectancy and social influence positively predicted behavioral intention. Based on these findings, this study offers recommendations on how to support the adoption of IVR for professional collaboration, facilitating the transfer of the reported benefits of IVR from the literature to practice.

Background Mood-behavioral symptoms, fatigue and pain are frequent among people with multiple sclerosis (pwMS). Music therapy (MT) is a non-pharmacological option for symptomatic treatment in neurological diseases. Objectives To assess effects of 6-week-outpatient MT on anxiety (primary outcome) as well as: depression, fatigue, pain and body perception (secondary outcomes) in pwMS. Methods We randomized pwMS 1:1 to music therapy group (MTG) and control group (CG). Both had one 45-min session per week, MTG with a monochord, CG without music. A blinded rater assessed endpoints at baseline and week 6 with standardized questionnaires (e.g. hospital anxiety and depression scale, HADS) and quantitative sensory testing (QST). Immediate session effects were also assessed. The analysis included linear mixed models, adjusted for pwMS's characteristics and baseline scores. Results Fifty-seven pwMS (age: 50.1 ± 12.4 years, 47 women, MTG: n = 30, CG: n = 27) were included. In MTG, anxiety levels (HADS) did not differ from CG at week 6 (p = 0.109). Among secondary outcomes, psychosocial fatigue was reduced (p = 0.029), QST heat pain thresholds were higher (p = 0.024) and immediate subjective effects stronger in MTG (e.g. feeling balanced: p < 0.001, relaxed: p < 0.001, less pain: p < 0.001). Conclusion Despite no difference in anxiety, we observed effects of receptive MT on fatigue, pain and body perception.

This paper addresses the growing complexity of 6G Radio Resource Management (RRM) due to the integration of new multimedia services and sophisticated network functionalities. To handle complex RRM control issues, machine learning is recognized as a valuable tool for improving decision-making in scheduling. However, the effectiveness of intelligent scheduling methods can be limited without efficiently managing the large volumes of data produced by network measurements, such as Channel Quality Indicator (CQI) reports. To tackle this, we propose a low-complexity framework that compresses extensive CQI data into a compact, usable form for machine learning applications. The framework adopts a metaheuristic approach with simulated annealing to perform clustering and identify global CQI data centers. These centers are then used to train a Radial Basis Function Network (RBFN), enabling efficient classification of CQI data into distinct patterns. These patterns inform reinforcement learning-based scheduling decisions, which, in turn, achieve significant gains, up to 20% improvement, in fairness performance when scheduling multimedia consumers.

PurposeOur aim was to develop new evidence-based and consensus-based recommendations for imaging strategies in patients with multiple and/or severe injuries in the resuscitation room. This guideline topic is part of the 2022 update of the German Guideline on the Treatment of Patients with Multiple and/or Severe Injuries.MethodsMEDLINE and Embase were systematically searched to August 2021. Inclusion criteria: patients with multiple and/or severe injuries in the resuscitation room, randomised controlled trials, prospective cohort studies, cross-sectional studies, and comparative registry studies; comparison of interventions for imaging strategies; patient-relevant clinical outcomes such as diagnostic test accuracy and mortality. Further literature reports were obtained from clinical experts. We considered patient-relevant clinical outcomes such as diagnostic test accuracy and mortality. Risk of bias was assessed using NICE 2012 checklists. The evidence was synthesised narratively, and expert consensus was used to develop recommendations and determine their strength.ResultsTwenty-one studies with a total of 55,227 patients were identified. There were studies with low risk (n = 2), high risk (n = 5) and unclear risk of bias (n = 14). Relevant topics were sonographic imaging of the chest and abdomen (n = 8 studies), X-ray of the chest (n = 1), indications for whole-body computed tomography (n = 6), CT scanner location (n = 1), whole-body computed tomography in haemodynamically unstable patients (n = 3), and prehospital sonography (n = 2). There were studies with low risk (n = 2), high risk (n = 5) and unclear risk of bias (n = 14). One new recommendation was developed, six were modified. All achieved strong consensus.ConclusionWhile extended focused assessment with sonography for trauma should be performed for diagnostic purposes after blunt and/or penetrating thoracic and/or abdominal trauma as part of the primary survey in the resuscitation room, whole-body computed tomography (WBCT) gains highest importance as part of the diagnostic procedures for severely injured patients. WBCT with a trauma-specific protocol must be performed in a timely manner if the patient does not require an immediate intervention. Magnetic resonance imaging can be indicated as a further primary diagnostic tool for specific conditions. Two studies were judged to be of low risk of bias in all domains. The risk of selection bias was high in two studies and unclear in seven studies.

Due to the relative motion of the solar system through its local interstellar environment, interstellar dust can enter the heliosphere. The dust trajectories are influenced by gravitational forces, solar radiation pressure force and by the solar wind through the Lorentz force on charged dust grains. Finally, these interstellar micro-particles can be measured in the solar system by in situ cosmic dust instruments. One of the most extensive in situ interstellar dust databases stems from the Ulysses dust detector. However, distinguishing interplanetary from interstellar dust is a major challenge and the selection criteria may influence the conclusions made based on the interstellar dust dataset. In this work, we elaborate on how the selection criteria and mass determination methods influence the inferred ISD population bulk properties like the dust particle flux, flow direction, and the gas-to-dust mass ratio in the interstellar medium. We examine the largest dust grains that were selected as "interstellar" and we illustrate a methodology to infer the filtering at the heliosphere's outer regions (heliosheath) using in situ data in the solar system, simulations of dust transport in the heliosphere inside of the termination shock, and assumptions on the initial dust size distribution from astronomical observations. We conclude with lessons learned for future detectors, in particular the necessity to use velocity grids and large detector surface areas. 

Procrastination and other forms of irrational delay are widespread among university students, leading to an array of potential negative consequences. While the reasons for this type of behavior are manifold and many facilitating factors have been identified, which of these factors are able to predict dilatory behavior in online/distance education has received comparatively little attention in the literature so far. In this study, we intended to compare the performance of two sets of objective predictors of delay, namely contextual variables based on characteristics of the assignment, and behavioral variables based on log data. Using historical data drawn from our university's learning management system, we calculated Bayesian multilevel models. The strongest and most consistent predictors of dilatory behavior turned out to be interval between the first click on the assignment and its deadline, the interval between the start of a block and the first click on the assignment, the number of clicks on the assignment, and the deadline type. The combination of both sets of predictors slightly improved the model's performance.

The forthcoming wireless network is expected to support a wide range of applications, from supporting autonomous vehicles to massive Internet of Things (IoT) deployments. However, the coexistence of diverse applications under a unified framework presents several challenges, including seamless resource allocation, latency management, and systemwide optimization. Considering these requirements, this paper introduces WIND (Wireless Intelligent Network Digital Twin), a self-adaptive, self-regulating, and self-monitoring framework that integrates Federated Learning (FL) and multi-layer digital twins to optimize wireless networks. Unlike traditional Digital Twin (DT) models, the proposed framework extends beyond network modeling, incorporating both communication infrastructure and application-layer DTs to create a unified, intelligent, and context-aware wireless ecosystem. Besides, WIND utilizes local Machine Learning (ML) models at the edge node to handle low-latency resource allocation. At the same time, a global FL framework ensures long-term network optimization without centralized data collection. This hierarchical approach enables dynamic adaptation to traffic conditions, providing improved efficiency, security, and scalability. Moreover, the proposed framework is validated through a case study on federated reinforcement learning for radio resource management. Furthermore, the paper emphasizes the essential aspects, including the associated challenges, standardization efforts, and future directions opening the research in this domain.