Practical Applications Research Articles

Soil Health and Challenges to Sustainable Soil Management in Denmark: Stakeholder Perceptions

ABSTRACTMaintaining soil health is essential to ensuring an adequate food supply and preserving the environment. Insights from multiple stakeholder inquiries can provide a more nuanced understanding of conditions for soil health to support the adoption of sustainable soil management practices to meet national and regional goals. This article aimed to gather insights from stakeholders' perceptions of soil health and the state of soil knowledge in Denmark. Seven stakeholder inquiries, collected through a series of recent European soil research activities, were synthesised to identify perceptions of soil health, concerns about soil challenges and needs for harmonising production and use of knowledge related to sustainable soil management. Each inquiry was analysed individually, and common themes were discussed across topics. The data showed that the ecosystem service framework related to soil health was not evenly familiar across stakeholder categories. Identified concerns for soil health included climate change effects, knowledge transfer, economic pressure/risk, and soil challenges. The top ‘priority’ soil challenges identified were improving soil organic matter/peat soil, avoiding soil compaction and improving nutrient retention/use efficiency. Soil knowledge gaps related to soil challenges included: multidisciplinary and interdisciplinary long‐term experiments, site‐specific measures, science–policy–stakeholder interactions and knowledge feasibility at the farm level. Identified barriers preventing wider adoption of sustainable management practices included farm applicability of practices, farmers' engagement, knowledge sharing and lack of regulations for challenges such as soil compaction. Farmers were also concerned about the time and effort required to learn new practices, especially as it relates to their work/life balance and the challenge of implementing sustainable practices on commercial farms while ensuring profitability. These concerns might partially arise from knowledge gaps between stakeholder categories. Raising awareness of sustainable practices and addressing current and future risks such as pests and weather extremes are crucial for policy and stakeholder engagement.

Process development of guava leaves with alkali in removal of zinc ions from synthetic wastewater

BackgroundThis research investigation delves into the efficient removal of zinc ions from synthetic wastewater through the utilization of KOH treated guava leaves as an adsorbent. The study employs advanced analytical techniques including SEM, FTIR, XRD, and BET analysis to characterize the physicochemical properties of the treated guava leaves. The research extensively explores the impact of various experimental factors on the adsorption process, encompassing agitation time, adsorbent dosage, pH levels, and desorption, to identify optimal conditions for maximum Zn (II) ion removal. MethodsThe research methodology involves subjecting guava leaves treated with KOH to thorough analysis using SEM, FTIR, XRD, and BET techniques to gain insights into their physical and chemical attributes. The study systematically investigates the adsorption process by manipulating crucial parameters such as agitation time, adsorbent dosage, and pH levels. Different isotherm models including Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich are applied to comprehend the adsorption mechanism. Additionally, kinetic calculations are performed, leading to the determination that the pseudo-second-order model best fits the removal of zinc metal. Significant FindingsThe research uncovers essential findings regarding the efficacy of KOH treated guava leaves as an adsorbent for zinc ions. Optimum conditions are pH 3 are identified, an adsorbent dosage of 100 mg/L, and an equilibrium period of 20 min, resulting in peak efficiency of the adsorption process. The study reveals an adsorption capacity of 14.5 mg/g for the studied metal, emphasizing the potential of KOH treated guava leaves as an efficient adsorbent for zinc ions. The adsorption capacity is 5.1 mg/g for the untreated guava leaves. The findings underscore the feasibility of utilizing this eco-friendly approach for wastewater management, thereby contributing to environmental remediation and pollution control efforts. The research highlights the practical application of KOH treated guava leaves in wastewater treatment processes, advocating for further exploration and optimization of this sustainable solution to address heavy metal contamination and promote greener technologies for water purification and waste management.

Nanoencapsulation of Zataria multiflora Essential Oil Containing Linalool Reduced Antibiofilm Resistance against Multidrug-resistant Clinical Strains

Background: The rise in antimicrobial resistance, caused by the production of biofilms by bacteria, is a significant concern in the field of healthcare. Nanoemulsion technology presents itself as a viable alternative in the quest to circumvent antibiotic resistance in pathogenic bacteria. Objective: The aim of this research was to form a sustainable nanoemulsion from Z. multiflora, and evaluate its antibacterial and anti-biofilm activities against the clinical isolates of Pseudomonas aeruginosa, Proteus mirabilis, and Staphylococcus aureus. Materials and Methods: Bioactive compounds of the oil were identified using GC-MS. Zataria multiflora essential oil (ZMEO) nanoemulsion was formulated as a water-dispersible nanoemulsion with a diameter of 184.88 ± 1.18 nm. The antibacterial and antibiofilm activities of the essential oil in both pure and nanoemulsion forms were assessed against pathogenic bacteria causing hospital-acquired infections using minimal inhibitory concentrations (MICs) and the microtiter method, respectively. Results: The main constituents were found to be linalool (78.66 %), carvacrol (14.25 %), and α- pinene (4.53%). Neither ZMEO nor the emulsified ZMEO showed any antimicrobial activity. However, ZMEO exhibited a low inhibition of biofilm formation by P. mirabilis, S. aureus, and P. aeruginosa. The most promising finding was that when the emulsified ZMEO was present at a concentration of 750 μg/mL, it significantly reduced biofilm formation by the aforementioned bacteria to 39.68% ± 2.62, 56.54% ± 3.35, and 59.60% ± 2.88, respectively. This result suggests that ZMEO nanoemulsion has the potential to effectively disrupt persistent biofilms and enhance the penetration of antimicrobial agents into the biofilm matrix. Conclusion: In conclusion, the study provides evidence supporting the use of ZMEO nanoemulsion as a potential treatment option for combating biofilm-related infections caused by Pseudomonas aeruginosa, Proteus mirabilis, and Staphylococcus aureus. Further research is warranted to explore the practical application of the proposed essential oil in clinical settings.

Fully printed non-contact touch sensors based on GCN/PDMS composites: enabling over-the-bottom detection, 3D recognition, and wireless transmission

ABSTRACT The rapid advancement in intelligent bionics has elevated electronic skin to a pivotal component in bionic robots, enabling swift responses to diverse external stimuli. Combining wearable touch sensors with IoT technology lays the groundwork for achieving the versatile functionality of electronic skin. However, most current touch sensors rely on capacitive layer deformations induced by pressure, leading to changes in capacitance values. Unfortunately, sensors of this kind often face limitations in practical applications due to their uniform sensing capabilities. This study presents a novel approach by incorporating graphitic carbon nitride (GCN) into polydimethylsiloxane (PDMS) at a low concentration. Surprisingly, this blend of materials with higher dielectric constants yields composite films with lower dielectric constants, contrary to expectations. Unlike traditional capacitive sensors, our non-contact touch sensors exploit electric field interference between the object and the sensor’s edge, with enhanced effects from the low dielectric constant GCN/PDMS film. Consequently, we have fabricated touch sensor grids using an array configuration of dispensing printing techniques, facilitating fast response and ultra-low-limit contact detection with finger-to-device distances ranging from 5 to 100 mm. These sensors exhibit excellent resolution in recognizing 3D object shapes and accurately detecting positional motion. Moreover, they enable real-time monitoring of array data with signal transmission over a 4G network. In summary, our proposed approach for fabricating low dielectric constant thin films, as employed in non-contact touch sensors, opens new avenues for advancing electronic skin technology.

Sustainable Agricultural Transformation: Insights from the Emilia-Romagna Region

Today, Kazakhstan's agricultural industry remains one of the most risk-prone, heavily influenced by adverse weather conditions, volatile market prices, and insufficient infrastructure. The study aims to assess opportunities for enhancing Kazakhstan's agricultural sector's competitiveness through the adoption of smart technologies, drawing lessons from the Emilia-Romagna region in Italy. The research employs a combination of comparative analysis and case study methodologies to evaluate the applicability of international best practices in the context of Kazakhstan's agricultural development. The study relies on a variety of data sources, including government reports, international case studies, and performance metrics from both Kazakhstan and the Emilia Romagna Region. The findings demonstrate that integrating technologies such as IoT, robotics, and blockchain can improve productivity, sustainability, and market attractiveness in the agricultural sector. The study also highlights the need to enhance farmer and manager competencies and promote family businesses and cooperatives. By adopting these strategies, Kazakhstan could address the existing challenges in its agricultural sector, reduce vulnerabilities, and create a more robust agro-industrial ecosystem. Future research should focus on developing specific pilot programs that apply these integrated approaches at a regional level, as well as assessing the long-term impacts of digital and sustainable practices in enhancing the competitiveness of Kazakhstan's agricultural sector.

Cross-National Findings of Factors Affecting the Acceptance of AI-Based Sustainable Fintech

This study utilized mixed (qualitative and quantitative) methods to discover the current research trends for AI in sustainable Fintech and to validate a research model through empirical analysis. The primary purpose of this research is to explore the factors influencing the acceptance of AI tools within the sustainable Fintech industry through a cross-national perspective, identifying key benefit and sacrifice dimensions, along with sustainability considerations, that affect users’ intentions to adopt AI tools. Drawing on a bibliometric keywords approach, we first conducted an overall review of academic literature using Web of Science and VOSviewer (version 1.6.17), covering areas related to AI applications in Fintech and sustainable Fintech practices. Additionally, for a cross-national study, this study built and validated a conceptual framework on the intention to use AI tools by selecting subjects from Republic of Korea and China. As core theoretical premises of the conceptual framework, the study drew on the Value-Based Adoption Model (VAM) and the Technology Acceptance Model (TAM). Furthermore, we extended the TAM to embrace sustainable dimensions (perceived responsibility and perceived transparency). Overall, the study concludes that AI not only improves Fintech efficiency but also significantly contributes to sustainable development, suggesting collaboration between experts in AI, finance, sustainability, and other relevant fields for more research on AI integration with sustainable Fintech management. This research contributes to existing literature by highlighting the synergistic benefits of combining AI and sustainable Fintech and offers practical insights for industry practitioners and policymakers.

Mengintegrasikan Kecerdasan Emosional, Kecerdasan Spiritual Dan Kecerdasan Buatan Dalam Praktik-Praktik Manajemen Sumber Daya Manusia Pada Organisasi Bisnis

With the increasingly complex and dynamic business world, the integration of emotional intelligence (EQ), spiritual intelligence (SQ), and artificial intelligence (AI) into human resource management (HRM) practices offers a holistic approach to improving workplace effectiveness and well-being. Emotional intelligence plays a vital role in managing interpersonal relationships, communication, and conflict management, which are crucial for effective leadership and teamwork. Spiritual intelligence supports the search for meaning and purpose in work, creating an organizational culture that aligns with personal and collective values. Artificial intelligence, on the other hand, automates processes, provides data-driven analytics, and improves operational efficiency through cutting-edge technology. Integrating these three intelligences can create a more productive, innovative, and fulfilling work environment. This article examines the basic theories of each intelligence and explains how their application in HRM practices can yield significant benefits to organizations.

Scientific Translations of “Da-xue” in 19th Century Russia: A Comparative Analysis of Approaches of Nikita Bichurin and Dmitry Sivillov

This article presents a comparative analysis of two translations of the Confucian text “Da-xue,” undertaken by prominent Russian sinologists N. Ya. Bichurin and D. P. Sivillov in the 19th century. The aim of the study is to juxtapose the scholars’ approaches to translating and interpreting this classical Chinese philosophical text. The research material consists of manuscript translations and original Chinese canonical texts. The findings demonstrate a high level of scholarly accuracy in both translations, despite differences in methodology. Bichurin’s translation, based on a Qing dynasty educational resource, is distinguished by its rich commentary and cultural explanations. In contrast, Sivillov’s work, which relies on the original “Da-xue” text, is characterized by its conciseness and fidelity to the source. The analysis reveals divergent translation strategies employed by the scholars: Bichurin aimed for a more literal translation while preserving culturally specific elements, whereas Sivillov adapted the text for the Russian readership. Both translations significantly contributed to the development of scientific translation methodology for Chinese classics in 19th century Russia. The study underscores the value of these works for the understanding of Confucian philosophy and the history of Russian sinology. The results may find application in contemporary translation practices and in teaching the history of sinology.

Axial compressive performance of low-carbon high-strength recycled aggregate concrete

AbstractThis paper introduces a novel material, low-carbon high-strength recycled aggregate concrete (LCHRAC), developed by activating ground granulated blast-furnace slag (GGBS) and silica fume (SF) in an alkaline environment and integrating recycled aggregate. To evaluate its mechanical properties, uniaxial compressive tests were performed, systematically analyzing the effects of recycled concrete aggregate (RCA) substitution ratios, as well as the characteristic parameters of steel and polypropylene(PP) fibers, on LCHRAC’s mechanical behavior. The results indicated that compressive strength shows a gradual decline as the RCA substitution ratio increases, with a moderate reduction of 7.1% up to 50% replacement, and a more significant drop, retaining only 68.6% at 100% replacement. In contrast, the peak strain increases linearly, showing a 29% improvement at full replacement, while the toughness index exhibits a consistent upward trend, increasing by approximately 123% at a 100% replacement rate. Based on experimental data, empirical models were developed to predict the influence of key control variables on the compressive strength, peak strain, elastic modulus, and the uniaxial compressive stress–strain behavior of LCHRAC. Additionally, advanced characterization techniques, including X-ray diffraction (XRD), thermogravimetric analysis (TGA/DTG), scanning electron microscopy (SEM), and energy dispersive spectroscopy (EDS), were employed to elucidate the hydration mechanisms of the slag-silica fume composite system. These findings provide a comprehensive understanding of the mechanical performance and microstructural characteristics of LCHRAC, contributing to its potential application in sustainable construction practices.

Assessing the applicability of international AI practices in construction management: A systematic review with implications for Vietnam

This study systematically reviews 94 journal articles published between 2000 and 2024 to examine the application of artificial intelligence (AI) in construction management (CM). The review focuses on AI's role in managing quality, time, and cost, which are critical factors for project success. The findings reveal a growing trend in AI adoption, particularly in areas such as scheduling and quality control. However, the use of AI for on-site CM, especially in Vietnam, remains limited and underexplored. This research aims to assess international AI practices and their potential applicability to Vietnam, analyzing how these technologies could improve efficiency, accuracy, and safety in construction projects. The implications of these findings suggest that embracing AI technologies could lead to enhanced project outcomes in Vietnam's construction industry by optimizing quality management, improving scheduling efficiency, and better controlling costs.

FUNDAMENTALS OF MODERN METHODS OF RESEARCH OF GEOTECHNICAL AND HYDROGEOLOGICAL SYSTEMS

The article examines the development of a synergistic paradigm in the Earth sciences, in particular in geotechnics and engineering hydrogeology. It was determined that in the second half of the 20th century, as a result of changes in scientific methodology, there was a need to rethink linear approaches to the study of natural systems. Synergetics, as a science of interaction and self-organization of complex systems, became the basis for new scientific directions, in particular in the field of geosystems and hydrogeotechnical systems research. The work emphasizes the importance of introducing synergistic principles into geological and hydrogeological research, particularly in the context of dynamic processes, such as filtration in hydrogeosystems. It is shown that these processes, in particular in the regimes of hydrodynamic, hydrochemical and hydrothermal changes, have a non-linear nature and can be described by appropriate mathematical models. The work also covers the issue of the influence of climatic and anthropogenic factors on the state of groundwater and soil, which is important for the design of buildings and infrastructure in various regions. The mechanisms that affect the stability of hydrogeosystems, as well as their possible critical states, which can lead to turbulence or irreversible changes in the structures of water-bearing rocks, are considered separately. An important aspect is the human impact on these systems, which can be both directed and accidental, which emphasizes the need to consider synergistic effects to improve natural resource management and minimize negative environmental impacts. The article is aimed at highlighting the potential of synergy in the study of hydrogeotechnical systems and the need for its application in engineering and environmental protection practices. This information will be useful for engineers, geologists, hydrogeologists and students of relevant specialties, as well as for those engaged in research in the field of geotechnics and hydrogeology.

Artificial Intelligence and the Management of Viral Respiratory Infections inChildren

Viral respiratory infections in children are a major public health issue, with high incidence rates and a significant impact on healthcare systems. The application of artificial intelligence (AI) in the medical field offers substantial opportunities for early detection, accurate diagnosis, effective management, and prevention of these infections. Aim: This study aims to analyse the most effective AI-based approaches for managing viral respiratory infections in children, including its application in paediatric hospitals, telemedicine, and routine practices, while also identifying challenges associated with implementation. Methodology: A systematic literature review was conducted following the PRISMA guidelines. The search was performed across 10 major databases: De Gruyter, MDPI, Nature, PubMed, ScienceDirect, Elsevier, SpringerLink, Wiley Online Library, Taylor & Francis, and Frontiers, focusing on articles published between 2020 and 2024. Out of 46,900 scientific articles, 17 relevant studies were selected, including original research, meta-analyses, and systematic reviews.Results: AI has shown high efficiency in the early detection of symptoms, differential diagnosis between viral and bacterial infections, monitoring disease progression, and personalising treatments. Its use in telemedicine and family education has improved accessibility to care and raised awareness. Integration of AI in paediatric hospitals has reduced diagnostic time and optimised resources. However, large-scale implementation depends on collaboration between medical professionals and IT specialists. Conclusions: AI represents a promising solution for improving the management of viral respiratory infections in children. The development of standardised protocols and addressing ethical challenges are essential for the effective integration of this technology into paediatric practice.

Role and Diagnostic Significance of Apolipoprotein D in Selected Neurodegenerative Disorders.

The World Health Organization in 2021 ranked Alzheimer's disease and other dementias as the seventh leading cause of death globally. Neurodegenerative disorders are progressive, intractable, and often fatal diseases. Early diagnosis may allow patients to enjoy prolonged survival with attenuated symptomatology because of early intervention. Hence, further research on finding non-invasive biomarkers of neurodegenerative diseases is warranted. Apolipoprotein D (ApoD) is a glycoprotein involved in lipid metabolism, oxidative stress regulation, and inflammation. It is expressed in various body fluids and regions of the central nervous system. ApoD's roles in neuroprotection, lipid transport, and anti-inflammatory processes are crucial as far as the prevention of neurodegenerative pathologies is concerned. This review aims to summarize the background knowledge on ApoD, and it covers studies indexed in the PubMed, Scopus, and Web of Science databases. It discusses the evidence for the multifaceted roles of ApoD in the mechanisms and pathogenesis of multiple sclerosis, Alzheimer's disease, and Parkinson's disease. ApoD may be a specific, sensitive, easily obtained, cost-effective biomarker for neurodegenerative diseases and its applications in diagnostic practices, treatment strategies, and advancing neurodegenerative disorders' management.

Delphi study and qualitative analysis to establish domains, core competencies, and content in forensic nursing pedagogy

BACKGROUND: The American Nurses Association designated forensic nursing (FN) a specialty in 1995. Nursing schools subsequently adopted Lynch's theoretical framework. Inconsistent sub-specialty-focused graduate education emerged, creating difficult-to-compare early programs of study. AIMS: Update FN domains, descriptions of and context for FN practices, core competencies, and sub-competencies with content for graduate curricula, align FN with new AACN Essentials, and create a certification foundation. METHODS: Methods include the mini-Delphi design (Estimate-Talk-Estimate), with Nominal Group Techniques methods, Cognitive Task Analysis, and NVivo™ qualitative analysis of experts' virtual recordings. The FN sample (N=126) included educators-clinicians, who gathered face-to-face (2002-4, 2014) and virtually (2020-22). RESULTS: The 2002-04 results informed the 2014 meeting, validating major knowledge pillars (legal foundations, forensic science, and nursing science). Analysis of 2020-2022 data named key domains, descriptions of and context for practice, core competencies, sub-competencies, concepts, and content in forensic nurse science and practice. The research included qualitative analysis of participant conversations, naming key content for each pillar. CONCLUSIONS: The 20-year Delphi study transformed knowledge integration for competent future Generalist and Advanced Forensic Nurse education and practice. The study confirmed Lynch’s thesis, three pillars of knowledge, and aligned FN core competencies with AACN Essentials and other national nursing publications, naming common practices, essential for all FNs. The study’s derivation of scientific content, important for certifications, arose from qualitative analysis of expert conversations, implying that as FN science evolves, the diverse practice roles of the FN mature with a common and basic core for application in all FN practices.

Performance Analysis of Semi-refined Digital Forearm Modeling and Simplified Forearm Model in Electromagnetic Simulation

Objective: The objective of this study is to analyze the performance difference between simplified and digital models based on medical images. Methods: According to the characteristics of human anatomy, the finite element simulation software COMSOL Multiphysics 5.5 was employed to construct a simplified arm model using cylinders and a digital arm model based on Chinese digital human regarding electroacupuncture therapy as an example. A comparative analysis was then performed considering three aspects: mesh number, potential distribution, and resource consumption. Results: Through analysis, the digital arm model based on Chinese digital human requires significantly more mesh cells than the simplified arm model in mesh generation. Meanwhile, because the digital arm model based on the Chinese digital human fully expresses the nonuniformity of the tissue distribution in a real human body, its signal distribution in its interior is also relatively scattered, and the coupling potential slightly differs at the electrode vertex with the smallest change. In addition, the digital arm model has much higher resource consumption and computer hardware resource requirements compared with the simplified arm model. Conclusion: As a result, the digital model based on the Chinese digital human can more fully express the tissue distribution and electrical signal characteristics of a real human body. However, due to its high computational requirements, appropriate simplification can be selected to improve the computational efficiency of the model in practical applications.

Site-Invariant Meta-Modulation Learning for Multisite Autism Spectrum Disorders Diagnosis.

Large amounts of fMRI data are essential to building generalized predictive models for brain disease diagnosis. In order to conduct extensive data analysis, it is often necessary to gather data from multiple organizations. However, the site variation inherent in multisite resting-state functional magnetic resonance imaging (rs-fMRI) leads to unfavorable heterogeneity in data distribution, negatively impacting the identification of biomarkers and the diagnostic decision. Several existing methods have alleviated this shift of domain distribution (i.e., multisite problem). Statistical tuning schemes directly regress out site disparity factors from the data prior to model training. Such methods have a limitation in processing data each time through variance estimation according to the added site. In the model adjustment approaches, domain adaptation (DA) methods adjust the features or models of the source domain according to the target domain during model training. Thus, it is inevitable that it needs updating model parameters according to the samples of a target site, causing great limitations in practical applicability. Meanwhile, the approach of domain generalization (DG) aims to create a universal model that can be quickly adapted to multiple domains. In this study, we propose a novel framework for disease diagnosis that alleviates the multisite problem by adaptively calibrating site-specific features into site-invariant features. Specifically, it applies directly to samples from unseen sites without the need for fine-tuning. With a learning-to-learn strategy that learns how to calibrate the features under the various domain shift environments, our novel modulation mechanism extracts site-invariant features. In our experiments over the Autism Brain Imaging Data Exchange (ABIDE I and II) dataset, we validated the generalization ability of the proposed network by improving diagnostic accuracy in both seen and unseen multisite samples.

Attention-Based Multimodal tCNN for Classification of Steady-State Visual Evoked Potentials and Its Application to Gripper Control.

The classification problem for short time-window steady-state visual evoked potentials (SSVEPs) is important in practical applications because shorter time-window often means faster response speed. By combining the advantages of the local feature learning ability of convolutional neural network (CNN) and the feature importance distinguishing ability of attention mechanism, a novel network called AttentCNN is proposed to further improve the classification performance for short time-window SSVEP. Considering the frequency-domain features extracted from short time-window signals are not obvious, this network starts with the time-domain feature extraction module based on the filter bank (FB). The FB consists of four sixth-order Butterworth filters with different bandpass ranges. Then extracted multimodal features are aggregated together. The second major module is a set of residual squeeze and excitation blocks (RSEs) that has the ability to improve the quality of extracted features by learning the interdependence between features. The final major module is time-domain CNN (tCNN) that consists of four CNNs for further feature extraction and followed by a fully connected (FC) layer for output. Our designed networks are validated over two large public datasets, and necessary comparisons are given to verify the effectiveness and superiority of the proposed network. In the end, in order to demonstrate the application potential of the proposed strategy in the medical rehabilitation field, we design a novel five-finger bionic hand and connect it to our trained network to achieve the control of bionic hand by human brain signals directly. Our source codes are available on Github: https://github.com/JiannanChen/AggtCNN.git.

Iterative Deep Structural Graph Contrast Clustering for Multiview Raw Data.

Multiview clustering has attracted increasing attention to automatically divide instances into various groups without manual annotations. Traditional shadow methods discover the internal structure of data, while deep multiview clustering (DMVC) utilizes neural networks with clustering-friendly data embeddings. Although both of them achieve impressive performance in practical applications, we find that the former heavily relies on the quality of raw features, while the latter ignores the structure information of data. To address the above issue, we propose a novel method termed iterative deep structural graph contrast clustering (IDSGCC) for multiview raw data consisting of topology learning (TL), representation learning (RL), and graph structure contrastive learning to achieve better performance. The TL module aims to obtain a structured global graph with constraint structural information and then guides the RL to preserve the structural information. In the RL module, graph convolutional network (GCN) takes the global structural graph and raw features as inputs to aggregate the samples of the same cluster and keep the samples of different clusters away. Unlike previous methods performing contrastive learning at the representation level of the samples, in the graph contrastive learning module, we conduct contrastive learning at the graph structure level by imposing a regularization term on the similarity matrix. The credible neighbors of the samples are constructed as positive pairs through the credible graph, and other samples are constructed as negative pairs. The three modules promote each other and finally obtain clustering-friendly embedding. Also, we set up an iterative update mechanism to update the topology to obtain a more credible topology. Impressive clustering results are obtained through the iterative mechanism. Comparative experiments on eight multiview datasets show that our model outperforms the state-of-the-art traditional and deep clustering competitors.

Transdermal Drug Delivery System of Linagliptin Sustained-release Microparticle Gels: In vitro Characterization and In vivo Evaluation.

Linagliptin (LNG) exhibits poor bioavailability and numerous side effects, significantly limiting its use. Transdermal drug delivery systems (TDDS) offer a potential solution to overcome the first-pass effect and gastrointestinal reactions associated with oral formulations. The aim of this study was to develop LNG microparticle gels to enhance drug bioavailability and mitigate side effects. Linagliptin hyaluronic acid (LNG-HA) microparticles were prepared by spray drying method and their formulation was optimized via a one-factor method. The solubility and release were investigated using the slurry method. LNG-HA microparticle gels were prepared and optimised using in vitro transdermal permeation assay. The hypoglycaemic effect of the LNG-HA microparticle gel was examined on diabetic mice. The results indicated that the LNG-HA microparticle encapsulation rate was 84.46%. Carbomer was selected as the gel matrix for the microparticle gels. Compared to the oral API, the microparticle gel formulation demonstrated a distinct biphasic release pattern. In the first 30 minutes, only 43.56% of the drug was released, followed by a gradual release. This indicates that the formulation achieved a slow-release effect from a dual reservoir system. Furthermore, pharmacodynamic studies revealed a sustained hypoglycemic effect lasting for 48 hours with the LNG microparticle gel formulation. These findings signify that the LNG microparticle gel holds significant clinical value for providing sustained release and justifies its practical application.

New Scalable and Efficient Online Pairwise Learning Algorithm.

Pairwise learning is an important machine-learning topic with many practical applications. An online algorithm is the first choice for processing streaming data and is preferred for handling large-scale pairwise learning problems. However, existing online pairwise learning algorithms are not scalable and efficient enough for large-scale high-dimensional data, because they were designed based on singly stochastic gradients. To address this challenging problem, in this article, we propose a dynamic doubly stochastic gradient algorithm (D2SG) for online pairwise learning. Especially, only the time and space complexities of O (d) are needed for incorporating a new sample, where d is the dimensionality of data. This means that our D2SG is much faster and more scalable than the existing online pairwise learning algorithms while the statistical accuracy can be guaranteed through our rigorous theoretical analysis under standard assumptions. The experimental results on a variety of real-world datasets not only confirm the theoretical result of our new D2SG algorithm, but also show that D2SG has better efficiency and scalability than the existing online pairwise learning algorithms.