One-shot Federated Learning (OFL) has emerged as a promising paradigm, enabling global model training with minimal communication overhead. In OFL, the server model is usually distilled from an ensemble of pre-trained client models, while the ensemble also facilitates synthetic data generation for the knowledge distillation process. Prior works show that the performance of the final model is fundamentally tied to both the quality of the synthetic data and the ensemble. However, existing methods often optimize these two components separately, overlooking their interaction. To address this coupled optimization problem and provide a unified solution to the dual challenges of data and model heterogeneity inherent in OFL, we introduce Co-Boosting++, a novel OFL framework where synthetic data generation and ensemble construction mutually enhance each other in an iterative fashion. First, we fix the ensemble and generate hard samples in an adversarial manner. These samples are crucial for enhancing the robustness of knowledge transfer, as they challenge the model to generalize better, thereby improving quality of the synthetic data and subsequent distillation process. Second, leveraging these hard samples, we enhance the ensemble via a Mixture of Experts (MoE) mechanism. MoE allows dynamic adjustment of ensemble weights based on the generated hard samples, which enables the ensemble to better capture diverse and heterogeneous knowledge from client models. Furthermore, we extend Co-Boosting++ to support the simultaneous generation of multiple heterogeneous target models, enabling efficient adaptation to diverse device constraints. Extensive experiments on benchmark datasets demonstrate that Co-Boosting++ consistently outperforms state-of-the-art methods due to its coupled optimization of data and ensemble quality. Additionally, Co-Boosting++ is highly practical in real-world model market scenarios, requiring no local training modifications, additional transmissions, or restrictions on client model architectures. Our code is available at https://github.com/rong-dai/Co-Boosting-PP.

Background In dentistry, pharmacology is usually taught during early undergraduate training, whereas prescribing decisions are mainly required later during clinical practice. This temporal gap may hinder the transfer of pharmacological knowledge to patient care and contribute to low prescribing confidence among dental students. Educational strategies that provide concise, point-of-care support may help address this challenge. Methods A descriptive educational study using a cross-sectional survey design was conducted during undergraduate dental clinical rotations. A visual, card-based pharmacology resource based on microlearning and just-in-time learning principles ( Odontomecum in Cards ) was implemented in the clinical setting. An anonymous online survey explored students’ perceptions of the resource, including self-reported confidence in pharmacological prescribing before and after use, perceived usefulness, usability, acceptability, and qualitative feedback. Results A total of 100 undergraduate dental students (39 fourth-year and 61 fifth-year) completed the survey. Self-reported confidence in pharmacological prescribing was higher after use of the cards in the overall sample and in both academic years. Students reported high perceived usefulness and ease of use of the resource in the clinical setting, and most indicated that they would recommend it to peers. Qualitative feedback emphasized the practicality of the cards and suggested improvements mainly related to accessibility and format. Conclusion A visual, card-based pharmacology resource integrated into dental clinical training was well accepted by students and associated with increased perceived confidence in prescribing. These findings support the potential value of microlearning and just-in-time educational approaches in dental pharmacology education.

Having a support system that provides novice users with quick and practical assistance for navigating learning management systems (LMS) is essential. This research aims to develop an interactive, video-based online tutorial with engaging elements for LMS usage at the Afyon Kocetepe University and to examine users' experiences. Despite the availability of a help-support structure for navigating the LMS interface, instructors have faced difficulties in adapting to it. To address this, we created a new modular and flexible online tutorial enhanced with multimedia components to assist users in navigating the LMS.The study employed a formative research method, presenting the tutorial to real users for evaluation. The findings revealed that the average usability score of the tutorial was 76.8. Additionally, the average scores for knowledge retention and transfer testscores related to the interactive video content were 73.2. No significant differences were found in usability scores, retention, or transfer test scores based on gender or field of study. In conclusion, the online tutorial, featuring interactive video content, demonstrated usability scores well above average and had a positive impact on both conceptual and procedural learning. These findings will provide valuable insights for the literature and inform future research focused on designing and implementing similar systems.

Alongside the editorial control mechanisms of traditional media, such as fact-checking and source verification, medical journalism ensures the quality of medical knowledge transfer, similar to the peer-review process of professional journals. However, this role, which is central to ademocracy, is increasingly threatened by the rise of the internet and social media. Today, most news from science and medicine reaches citizens as "end consumers" of information directly, i.e., without intermediary quality control according to established journalistic standards. Deteriorating working conditions have forced many medical journalists to move to other occupational fields. The loss of quality control in medical knowledge transfer has become adanger to society. It has never been so difficult to distinguish between reliable information and misinformation or disinformation.

In photovoltaic remote sensing image segmentation tasks, fully supervised methods can achieve high accuracy. However, the high cost of pixel-level annotation significantly limits their scalability in large-scale scenarios. To overcome this annotation bottleneck, this paper proposes a zero-shot cross-modal segmentation framework based on the visual-language pre-trained foundation model (CLIP). This approach harnesses CLIP’s cross-modal knowledge transfer capabilities to achieve precise extraction of photovoltaic targets without requiring any downstream training. This paper first introduces the Layer-wise Augmented Residual Attention (LARA) mechanism to enhance fine-grained detail representation in the feature space. Subsequently, a Cross-modal Semantic Attribution Module (CMSA) is designed to generate precise activation maps by leveraging image-text alignment gradient information. Finally, the Confidence-Aware Refinement Strategy (CARS) replaces the conventional training-based denoising process, directly producing high-quality binary segmentation masks through adaptive thresholding. Comparative experiments were conducted to evaluate the proposed method against various baselines using several public datasets with varying resolutions in Jiangsu Province including Unmanned Aerial Vehicles imagery, Beijing-2, Gaofen-2, and a self-created Sentinel-2 imagery covering multiple countries. Notably, the proposed method achieved an IoU of 70.3% on the Gaofen-2 PV03 dataset with a spatial resolution of approximately 0.3 m and 50.8% on the self-created Sentinel-2 PV_Sentinel-2 dataset with a spatial resolution of 10 m. Experimental results demonstrate that our proposed approach maintains excellent cross-domain generalisation capabilities while reducing annotation costs, thereby providing an efficient and viable technical pathway for the automated monitoring of large-scale photovoltaic facilities.

Virtual communities of practice in health and social care: Early outcomes from a developmental evaluation in Quebec.

Abstract In constrained multi-objective optimization problems (CMOPs), the discontinuity of the target space and the fragmentation of the feasible solution space caused by complex constraints make the optimization algorithm face irreconcilable conflicts between convergence, diversity, and feasibility. To this end, this paper proposes a dual population co-evolution algorithm based on a dynamic Manhattan-Harmony hybrid distance. The algorithm constructs a main and auxiliary population with complementary structures: the main population focuses on deep search in the feasible domain, the auxiliary population conducts global exploration in the infeasible area, and introduces an evolutionary stage perception mechanism for differentiated environmental selection. In particular, the proposed dynamic Manhattan-Harmony hybrid distance can effectively characterize the convergence and diversity characteristics of individuals and guide the auxiliary population to adopt adaptive selection strategies at different stages. In addition, the algorithm draws on the theory of biological potential energy diffusion and designs a dynamic resource allocation mechanism that combines three types of potential energy: goal orientation, constraint recovery, and structural diversity, to achieve adaptive scheduling of offspring resources. Furthermore, the constructed bidirectional knowledge transfer channel realizes information sharing and co-evolution between the main and auxiliary populations. Experimental results on 33 standard test functions and 12 real-world problems show that HDCMO outperforms many existing representative constrained multi-objective evolutionary algorithms in terms of convergence, feasibility, and distribution balance, and has significant performance advantages and adaptability.

Jordan’s masonry archaeology across limestone, sandstone, and basalt faces escalating threats from a disconnect between conservation and architectural education. Though Jordanian archaeology has evolved into a multidisciplinary field, architecture curricula prioritize technical training over the engineering complexities of endangered sites. This study argues that engaging future architects with ancient engineering as recoverable technical knowledge, rather than as objects for specialist intervention, is essential for cultivating advocates of archaeology. It aims to develop a constructivist framework for architectural archaeology that reorients education from mere intervention toward knowledge transfer through reverse engineering. A mixed-methods experiment with architecture students at Hashemite University engaged participants in deconstructing ancient techniques through digital documentation and structural simulation and then reconstructing this knowledge for contemporary applications. A four-domain framework operationalized object-laden epistemology (technical acquisition) and value-laden ontology (constructed advocacy). Findings revealed four transformative outcomes: science-making (recovering ancient engineering as legitimate knowledge); heritage-making (sites becoming living practice); temporality-making (past–present dialogue within presentism and futurism); and advocacy-making (students as ‘custodian-transmitters’ assuming professional stewardship). By integrating architectural archaeology into core curricula, this framework reaches future architects beyond specialized programs, addressing regional gaps in community support for endangered heritage while maintaining critical reflexivity regarding power and selection in archaeological discourse.

To explore the multilevel factors (barriers and facilitators) influencing maternal and child nutrition (MCN) from community perspectives using the Socioecological Model (SEM) as the guiding framework. An exploratory qualitative study combining semi-structured online interviews (Microsoft Teams) and photovoice method. Data were audio-recorded, translated, transcribed and analysed inductively using Braun and Clarke's thematic approach, with final themes mapped to the SEM levels. Two semi-urban communities in Ado-Odo Ota Local Government Area, Ogun State, Nigeria. Twenty-five participants (aged 25-75 years), including pregnant and lactating mothers, fathers, grandmothers, community leaders, and health workers, were recruited through purposive and snowball sampling. Despite basic nutrition awareness, maternal diets were predominantly carbohydrate-based while infant feeding was characterised by delayed breastfeeding initiation, prelacteal feeding, and suboptimal complementary feeding practices. At the interpersonal level, strong familial and community support was evident, however, grandmothers strongly influenced dietary taboos and health-seeking decisions. Organisational barriers, including poor facilities and negative staff attitudes, reduced trust in primary healthcare centres and encouraged reliance on private clinics and traditional providers. Community-level challenges such as poverty, rising food prices, unreliable water supply and poor waste disposal systems increased the risk of malnutrition. Nonetheless, some households demonstrated resilience through home gardening, bulk food purchasing and adequate hygiene practices. MCN is shaped by interconnected factors. Sustainable progress requires moving beyond individual knowledge transfer towards integrated, system-based actions that strengthen nutrition and health services, improve access to diverse foods, safe water and sanitation, and leverage existing community strengths.

Contrary to the argument that biological conceptualisations of mental distress serve as a smokescreen for the ills of capitalism in Western neoliberal societies, the findings show that the biological and individualised definitions of depression were present in the socialist Eastern Bloc. Taking the history of Polish psychiatry as a vantage point, this study introduces its historical development throughout late socialism and into the capitalist regime, revealing the continuity of biological conceptualisations of depression over time. It draws from archival research of the Polish Psychiatry journal, supplemented by interviews with psychiatrists from different generations. The study complicates the history of the European psychiatry during the Cold War period, by tracking knowledge transfers between East and West. It proves that the understanding of depression in late socialist Poland as an endogenous disease, an organic disorder independent of the social environment, was partly shaped by the Pavlovian paradigm and partly by Western neurobiological discoveries. Since the 1990s, with the dominance of American science, conceptualisations of depression as an endogenous disease have prevailed, albeit under a new wording of biological vulnerability. The study argues that the simultaneous impact of the Soviet and Western psychiatry in late socialist Poland was an instance of dual dependency resulting from the country’s geopolitical positioning. With the systemic transformation in the 1990s, a shift in the dependency pattern occurred, as Polish psychiatry clearly oriented towards the U.S. science.

Adaptive multi-teacher knowledge distillation framework with foundation models for medical image analysis.

Drawing on the extensive resources of the International Library of African Music (ILAM) in Makhanda, South Africa, this article examines the African Musical Arts Programme at Rhodes University as a case study in decolonial music education. Established in 2017, the programme integrates the knowledge systems of African musicians and teachers with ethnomusicological scholarship to address the structural and epistemological challenges of transformation within higher education. Through a narrative qualitative approach, the study draws on data gathered between 2017 and 2024 from students, lecturers, and community musicians involved in the programme. It critically analyses the programme’s evolution, highlighting both its successes and ongoing challenges. The findings show how Indigenous Knowledge Systems (IKS) and community-based participation have been central to curricular design, teaching strategies, and assessment models. The programme demonstrates how embodied learning, participatory and performance-based pedagogies, and transgressive approaches can contribute to more sustainable and inclusive music education practices. At the same time, it reveals persistent institutional constraints related to the recognition of community knowledge holders, the development of African theoretical frameworks, and the implementation of non-traditional assessment methods. By situating the Rhodes University case within broader continental debates on sustainability, revitalisation, and decolonial pedagogy, the article contributes to the themes of the Mapping Africa’s Musical Identities special issue. It illustrates how locally rooted and globally relevant African musical arts education can challenge Eurocentric models while advancing the preservation and creative renewal of African musical traditions. In doing so, it offers a framework for sustainable knowledge transfer that supports the ongoing transformation of African music education in the 21st century.

The International Community Service Program themed “Integration of Social Media Algorithms in Optimizing the Use of Digital Marketing for Micro and Small Enterprises (MSEs) in Indonesia and Malaysia” is a collaborative initiative between Universitas Jambi (UNJA) and Universiti Poly-Tech Malaysia (UPTM). This program aims to enhance the digital literacy capacity of micro-entrepreneurs through a comprehensive understanding of social media algorithms and the application of creative content strategies rooted in local wisdom and culture. The implementation methods include interactive training sessions, thematic workshops, intensive mentoring, and cross-country benchmarking to strengthen knowledge transfer and the adoption of best practices. The results demonstrate a significant improvement in participants’ digital competencies, reflected in their enhanced ability to optimize social media algorithms for product marketing, the establishment of a cross-border digitalpreneur community between Indonesia and Malaysia, and the development of adaptive digital marketing training modules and standard operating procedures (SOPs). The sustainability impact of this program not only increases the competitiveness and resilience of MSEs but also strengthens international academic collaboration between the two universities. Furthermore, this initiative contributes to achieving the universities’ Key Performance Indicators (KPI) and supports the development of an inclusive and sustainable digital economic ecosystem within the ASEAN region.

Decomposition based curriculum-style self-training for source-free universal domain adaptation in computational pathology.

OVERVIEW: The fourth-generation university (4GU) represents a fundamental shift in how universities engage with innovation ecosystems. While entrepreneurial universities emphasize commercialization and direct economic engagement, 4GUs explicitly organize their teaching, research, and valorization activities around societal transformation missions while orchestrating regional innovation ecosystems. For R&D managers, this transition creates new strategic opportunities: Small and medium-sized enterprises (SMEs) gain access to research infrastructure and collaborative networks otherwise beyond their reach, while large corporations can externalize exploratory research and participate in system-level solutions to grand challenges. We develop a working definition of the 4GU and demonstrate its practical implications through Eindhoven University of Technology’s evolution within the Brainport ecosystem, showing how this model creates value for R&D managers through ecosystem participation rather than bilateral knowledge transfer. We provide a staged implementation framework that guides firms from ecosystem assessment to co-orchestration, supported by multitier performance metrics that balance traditional innovation outputs with ecosystem development indicators. This article contributes to innovation management practice by reframing university–industry collaboration as ecosystem development—essential for addressing the complex, interdependent challenges that define contemporary innovation.

Dual-channel self-supervised multi-task learning for spectral detection of soluble solids content and firmness in Korla fragrant pears.

Anomaly detection on tabular data is widely used in fraud detection, predictive maintenance, and medical screening. While heterogeneous ensembles combining multiple detection paradigms achieve strong performance, their computational cost limits deployment in latency-sensitive or resource-constrained environments. We propose KD-AnomalyNet, a teacher–student framework that distills anomaly knowledge from a high-capacity ensemble into a lightweight neural model for efficient inference. Beyond performance replication, we study how anomaly representations transfer during distillation. To this end, we introduce a noise perturbation analysis that serves as a diagnostic probe for representation stability without introducing additional trainable components. Experiments on ten benchmark datasets show that the distilled model preserves up to 98.5% of the teacher’s AUC-ROC on the nine capacity-sufficient datasets (84.7% mean retention across all ten datasets) while achieving 26–181× inference speedups. Our analysis reveals which forms of anomaly knowledge transfer reliably—global outliers (78% transfer) and isolation-based detection (88% retention)—and which degrade under compression—local outliers (20% transfer) and neighborhood-based detection (76% retention)—providing practical guidance for deploying distilled anomaly detectors.

In the era of digital transformation and VUCA (volatility, uncertainty, complexity, ambiguity) environments, continuous organizational learning has become a strategic necessity to remain viable. While traditional classroom training was the standard, e-learning is increasingly adopted for its efficiency and flexibility in human resource development. However, doubts remain regarding the comparative effectiveness of digital versus face-to-face formats, specifically concerning dropout rates and social isolation. This study investigates the specific advantages of each format and the critical role of organizational learning culture in ensuring success. Based on a systematic literature review, this paper analyses theoretical foundations and practical applications to assess both learning modalities. It conducts a comparative assessment to determine their suitability for different organizational contexts and skill requirements. The analysis reveals that while e-learning excels in flexibility and factual knowledge transfer, face-to-face training is superior for social interaction and developing practical skills. Effectiveness depends less on the format itself and more on the didactic design and alignment with the organizational learning culture. These findings imply that organizations should move beyond an either/or approach towards blended learning strategies. To remain viable in the long term, they must promote Deutero learning (the ability to learn how to learn) within their cultural framework.

Augmented and virtual reality training versus conventional teaching methods for geriatric oral care practices in nursing assistants: A randomized controlled trial.

Farmers for Climate Solutions started the Farm Resilience Mentorship (FaRM) Program in 2021 as a peer-to-peer, regionally adaptive, education and knowledge sharing program. The program is designed to support farmers and ranchers in adopting low emissions, high resilience Beneficial Management Practices (BMPs), specifically: improved nitrogen management, cover cropping, and rotational grazing.<br/>With partner organizations across the country, FaRM supports in-person and virtual learning events, with farmers and ranchers, agrologists, agronomists, Certified Crop Advisers and agricultural researchers as knowledge sharers. It also includes the FaRM Learning Hub where farmers, ranchers and agronomists can take online self-directed courses to build their knowledge on how to adapt BMPs to their farm or ranch.<br/>With an adaptive approach to knowledge mobilization and a desire to support behaviour change, FaRM has developed a rigorous monitoring, evaluation and learning framework that has allowed FaRM to pivot to serve the needs of regional communities while still achieving program objectives. Of the farmers and ranchers responding to the 2024 FaRM Impact Survey, 89% indicated that they had implemented, expanded, and/or enhanced a BMP to some extent as a result of participating in the FaRM Program, with an additional 11% of farmer/rancher respondents indicating that they had yet to undertake a BMP, but intended to. However, despite this strong uptake, cross-country sector consultations have noted that without local, regional data, producers and agronomists are hesitant to implement practices. In 2025, in response to this, and to build engagement with agronomists, FaRM launched Phase 2, which focuses in part on increasing regional support and knowledge transfer by engaging agricultural professionals, in partnership with farmers and ranchers, to run field trials to overcome regional barriers to BMP adoption and to troubleshooting BMP adoption through practical, hands-on experience.