Search Results (372)

Epistemology remains tacitly anthropocentric: it treats knowledge as something produced and validated through human cognitive capacities such as understanding, intuition, and transparent justification. Yet contemporary science and artificial intelligence increasingly depend on non-human systems that generate mathematically valid results, empirically successful models, and operationally reliable inferences that no human can fully survey or interpret. This article develops xenoepistemics, a structural theory of non-anthropocentric knowledge. The central claim is that epistemic evaluation must be reformulated in terms of system-level properties—reliability, robustness, counterfactual sensitivity, and domain transfer—rather than mentalistic notions such as belief or understanding. I offer (i) a definition of xenoepistemic systems as systems that track structure in a target domain without requiring human-style semantic access; (ii) a minimal account of epistemic agency without minds that avoids trivialization; and (iii) a non-circular trust framework that distinguishes empirical success from epistemic legitimacy using independent validation regimes. This paper addresses a reflexive worry—that a human-authored theory cannot dethrone human epistemology—by separating standpoint from object: xenoepistemics is articulated by humans but is not about human cognition. I discuss the pragmatic value of xenoepistemic knowledge production, the limits of independent verification for opaque systems, domain-relative thresholds for xenoepistemic authority, and the problem of constitutionally human-inaccessible knowledge. Finally, I diagnose and formalize the Marcusian regress paradox: recurrent goalpost-shifting, whereby every machine competence is reclassified as irrelevant once achieved. Xenoepistemics reframes this debate by treating non-human knowledge as a present reality requiring new norms, not as a future curiosity. Full article

This commentary critically evaluates the translational relevance of a recent study investigating the effects of ozonated saline solution (O₃SS) on microglial and endothelial cell models. While the original research proposes potential antioxidant and anti-inflammatory benefits of low-dose ozone exposure, we identify significant methodological and conceptual flaws that undermine its conclusions. Key concerns include the unjustified assumption that ozone behaves similarly in microwell cultures and clinical infusion settings, despite known physicochemical differences affecting ozone stability and reactivity. The use of immortalized BV2 and HUVEC cells, which lack the complexity of in vivo systems, further limits the study’s applicability. The absence of accurate ozone quantification, proper controls, protein-level validation, and kinetic modeling exacerbates these weaknesses. Our analysis also demonstrates, through differential equation modeling, that ozone rapidly decays in saline solutions, making systemic delivery via infusion chemically implausible as a therapeutic approach. Moreover, the extrapolation of in vitro gene expression data to systemic therapeutic claims lacks scientific justification. We conclude that while the observed cellular responses in vitro are of academic interest, they do not support the efficacy or safety of O₃SS in clinical settings. A more rigorous approach is necessary to substantiate the biomedical potential of ozonated solutions. Full article

Recommender systems are central to helping users navigate the rapidly expanding entertainment ecosystem, yet achieving strong personalization with limited feedback while maintaining interpretability remains difficult, particularly under cold-start conditions and heterogeneous item metadata. This work presents an end-to-end hybrid recommendation framework that unifies a Large Language Model (LLM) with Graph Attention Network (GAT)-based collaborative filtering to improve both ranking accuracy and explanation quality across movies, books, and music. LLM-based agents first transform raw metadata such as titles, genres, descriptions, and auxiliary attributes into semantically grounded user and item profiles, which are embedded and used as initial node features in a user–item bipartite graph processed by a GAT-based recommender. Model optimization relies on a hybrid objective combining Bayesian Personalized Ranking, cosine-similarity regularization, and robust negative sampling to better align semantic and collaborative signals. Finally, in the post-processing stage, an LLM-based agent re-ranks the GAT outputs using a proposed Hybrid Confidence-Weighted Binary Search Tree, and another LLM-based agent that produces natural-language justifications tailored to each user. Experiments on diverse benchmark datasets and extensive ablations demonstrate that the proposed methodology increases precision, recall, NDCG, and MAP across various values of K. In addition, the post processing step is especially effective in cold-start scenarios, consistently strengthening recommendation metrics and enhancing transparency at smaller values of K. Overall, integrating LLM-enriched representations with attention-based graph modeling enables more accurate and explainable entertainment recommendations. Full article

Depression is notable for high rates of disability. The medical model typically characterizes depression as a physiological dysfunction or psychological disorder. However, both views fail to appreciate the phenomenology of depressed experience. Drawing on the existential phenomenology of Merleau-Ponty, this article contends that the lived experience of chronic depression is marked by a disturbance between the body-subject and the world. More specifically, the experience of depression is characterized by alienation from the world, self and others. While anti-depressants have long been the first line of treatment of depression, many governments subsidize cognitive behavioral therapy (CBT) as an adjunct treatment. CBT is said to be the gold standard psychotherapeutic treatment given that it is evidence-based, cost-effective, and short in duration. However, not only are these justifications questionable, but the theoretical underpinnings of CBT have ideological significance. Rather than approaching depressed persons as body-subjects, CBT casts service users as neo-liberal subjects, insofar as depression is characterized as disordered thinking that is independent of a person’s situated life. The emphasis on quickly returning people to work to reduce strain on welfare systems, while a valid economic concern, is not a valid therapeutic concern. The limited choice of subsidized psychotherapeutic options fails to recognize that depression is a heterogenous phenomenon, meaning that the CBT model of disordered thinking is not necessarily representative of the way in which depression manifests. Full article

Despite recent advances in artificial intelligence for agriculture, reliable crop recommendation remains constrained by limited access to soil diagnostics, insufficient integration of environmental context, and the absence of transparent, quantitative evaluation frameworks. This study addresses the research question: How can we integrate multiple indicators to generate accurate, explainable, and context-sensitive crop recommendations? To this end, we propose a multimodal decision-support framework that combines image-based soil texture classification with geospatial, and climatic information. A convolutional neural network was trained on a curated dataset of 3250 soil images aggregated from four publicly available sources, covering four primary soil texture classes, alongside tabular soil and nutrient data. The model was evaluated using 5-fold stratified cross-validation, achieving an average classification accuracy of 99.30% (standard deviation ≈ 0.66), and was further validated on an independent hold-out test set to assess generalization performance. To enhance practical applicability, the framework incorporates elevation, rainfall, temperature, and major soil nutrients, and employs a large language model to generate user-oriented, interpretable justifications for each recommendation. Crop recommendations were quantitatively evaluated using a novel Agronomic Suitability Score (ASS), which measures alignment across soil compatibility, climatic suitability, seasonal alignment, and elevation tolerance. Across six geographically diverse case studies, the framework achieved mean ASS values ranging from 3.76 to 4.96, with five regions exceeding 4.45, demonstrating strong agronomic validity, robustness, and scalability. A Streamlit-based application further illustrates the system’s ability to deliver accessible, location-aware, and explainable agronomic guidance. The results indicate that the proposed approach constitutes a scalable decision-support tool with significant potential for sustainable agriculture and food security initiatives. Full article

To enhance the early-age properties of steel slag cement mortar and promote the resource utilization of metallurgical solid waste, in this study, nano-SiO₂ (KH-NS) was modified using a KH550 silane coupling agent. The hydration kinetics and microstructure evolution were systematically analyzed by means of a macroscopic performance test (setting time and compressive strength) and multi-scale microscopic characterization (characterized by Fourier Transform Infrared Spectroscopy, Scanning Electron Microscopy, X-ray Diffraction, Thermogravimetry-Differential Thermal Analysis, and isothermal calorimetry). The influence mechanism of its content on the early performance of the steel slag cement system was systematically studied. Research findings indicate that at a given dosage, increasing the proportion of KH-NS results in a shorter setting time for steel slag mortar. When the KH-NS dosage reaches 1.5%, the initial and final setting times of steel slag mortar decrease by 24.21% and 21.20%, respectively. The addition of KH-NS effectively enhances the compressive strength of mortar, with a particularly pronounced effect on early strength prior to 14 h of curing. At a KH-NS dosage of 1.5%, the onset of the accelerated phase of hydration heat release in steel slag cement mortar is advanced by 2.5 h. Mechanistic studies indicate that KH-NS accelerates cement hydration by promoting C₃S dissolution and C-S-H gel nucleation through interactions between surface silanol groups (Si-OH) and amino groups (-NH₂). Furthermore, KH-NS refines the pore structure via a micro-aggregate filling effect, reducing the number of harmful pores and improving the pore size distribution. KH-NS continuously consumes Ca(OH)₂ through pozzolanic reactions to generate C-S-H, with its reactivity increasing with higher dosage. Research confirms that KH-NS significantly enhances the early strength and density of steel slag mortar, providing both theoretical justification and technical support for developing low-carbon building materials based on solid waste with high dosage. Full article

Non-parametric information geometry has long faced an “intractability barrier”: in the infinite-dimensional setting, the Fisher–Rao metric is a weak Riemannian metric functional that lacks a bounded inverse, rendering classical optimization and estimation techniques computationally inaccessible. This paper resolves this barrier by building the statistical manifold on the Orlicz space $L_0^{\mathsf{\Phi }}\left(P_f\right)$ (the Pistone–Sempi manifold), which provides the necessary exponential integrability for score functions and a rigorous Fréchet differentiability for the Kullback–Leibler divergence. We introduce a novel Structural Decomposition of the Tangent Space ($T_{f}M=S\oplus S^{\perp }$), where the infinite-dimensional space is split into a finite-dimensional covariate subspace (S)—representing the observable system—and its orthogonal complement ($S^{\perp }$). Through this decomposition, we derive the Covariate Fisher Information Matrix (cFIM), denoted as $G_f$, which acts as the computable “Hilbertian slice” of the otherwise intractable metric functional. Key theoretical contributions include proving the Trace Theorem ($H_G\left(f\right)=\mathrm{Tr}\left(G_f\right)$) to identify G-entropy as a fundamental geometric invariant; demonstrating the Geometric Invariance of the Covariate Fisher Information Matrix (cFIM) as a covariant $(0,2)$-tensor under reparameterization; establishing the cFIM as the local Hessian of the KL-divergence; and characterizing the Efficiency Standard through a generalized Cramer–Rao Lower Bound for semi-parametric inference within the Orlicz manifold. Furthermore, we demonstrate that this framework provides a formal mathematical justification for the Manifold Hypothesis, as the structural decomposition naturally identifies the low-dimensional subspace where information is concentrated. By shifting the focus from the intractable global manifold to the tractable covariate geometry, this framework proves that statistical information is not a property of data alone, but an active geometric interaction between the environment (data), the system (covariate subspace), and the mechanism (Fisher–Rao connection). Full article

The accelerated expansion of advanced technologies—particularly artificial intelligence, intelligent systems, and interactive digital environments—is influencing contemporary media ecosystems and contributing to changes in educational practices. This study provides a systematic and descriptive bibliometric mapping of recent scientific production on artificial intelligence in education, algorithmic ethics, and digital inequality. A total of 229 Scopus-indexed documents published between 2021 and 2026 were analyzed using Biblioshiny and VOSviewer to examine publication patterns, influential authors and sources, and the conceptual structure of the field. Results indicate a marked increase in research output since 2024, with an annual growth rate of 47.58%, an average of 8.68 citations per document, and an international co-authorship rate of 24.45%. These indicators reflect an expanding and increasingly collaborative research landscape, accompanied by a diversification of thematic priorities within the field. The analysis identifies five thematic clusters: (1) the technical foundations of AI and digital transformation; (2) intelligent and immersive learning environments; (3) personalized and adaptive learning systems; (4) AI literacy and pedagogical integration; and (5) ethical considerations, including algorithmic bias and educational robotics. The findings highlight the need for explicit justification of database selection, strengthened critical AI literacy, and context-sensitive strategies that address disparities in access, skills, and institutional capacity. Overall, this study offers a coherent overview of a research area that is currently expanding and undergoing conceptual reorganization, providing evidence-informed insights for future research, policy development, and the design of equitable AI-driven educational technologies. Full article

We describe AI agents as stochastic dynamical systems and frame the problem of learning to reason as in transductive inference: Rather than approximating the distribution of past data as in classical induction, the objective is to capture its algorithmic structure so as to reduce the time needed to solve new tasks. In this view, information from past experience serves not only to reduce a model’s uncertainty, as in Shannon’s classical theory, but to reduce the computational effort required to find solutions to unforeseen tasks. Working in the verifiable setting, where a checker or reward function is available, we establish three main results. First, we show that the optimal speed-up for a new task is tightly related to the algorithmic information it shares with the training data, yielding a theoretical justification for the power-law scaling empirically observed in reasoning models. Second, while the compression view of learning, rooted in Occam’s Razor, favors simplicity, we show that transductive inference yields its greatest benefits precisely when the data-generating mechanism is most complex. Third, we identify a possible failure mode of naïve scaling: in the limit of unbounded model size and computing, models with access to a reward signal can behave as savants, brute-forcing solutions without acquiring transferable reasoning strategies. Accordingly, we argue that a critical quantity to optimize when scaling reasoning models is time, the role of which in learning has remained largely unexplored. Full article

Hydroelectric power plants are renewable electricity generation assets that require high availability and reliability in their operation and maintenance. To justify improvement actions (modernization and investments), it is necessary to analyze the operation of the plant, the maintenance plan being implemented, and, naturally, the incidents and breakdowns that affect this asset. This paper presents research on hydroelectric power plant maintenance based on the development of a database of incidents and failures of such plants, considering the methodology of failure modes, effects and criticality analysis (FMECA) as well as the reliability-centered maintenance (RCM) methodology of the initial maintenance plan of a standard hydroelectric power plant. Different maintenance standards and analysis standards (IATF criticality of failure modes, UNE 13306, ISO 14224, etc.) were considered. The results reveal different improvement and optimization actions based on the current technological development, which can be applied to hydroelectric generation (Innovation 4.0), as well as actions to optimize the initial maintenance plan based on Maintenance 4.0. The technical justification for such improvements in hydropower generation highlights a key area of development in the expansion of renewable energies worldwide. Hydropower generation assets have contributed renewable energy to the system for many years; however, they now require redesign in their operation and maintenance. Full article

Background/Objectives: Dichorionic triamniotic (DCTA) triplet pregnancies are associated with increased rates of placenta-specific complications primarily attributed to vascular anastomoses in the monochorionic (MC) pair. Selective fetal reduction to twins (of one of the MC pair) is a complex and not a widely available procedure. Multifetal reduction (MFR) to singleton pregnancy can reduce adverse pregnancy outcomes but is controversial due to medico-legal and socio-ethical issues. The aim of this study is to identify the rate of miscarriage < 24 weeks or preterm birth < 34 weeks following MFR to singleton pregnancy in DCTA triplets and compare the results with expectant management. Methods: This systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and registered in the Prospective Register of Systematic Reviews System (ID: CRD42023422585). Results: Overall, from 21 citations of relevance, 6 studies with a total of 548 DCTA triplet pregnancies fulfilled the inclusion/exclusion criteria. In comparison with expectant management (n = 336), meta-analysis demonstrated that MFR to singleton pregnancy (n = 212) was associated with a lower rate (9.4% vs. 48.5%) of preterm birth (RR = 0.19, 95%CI 0.07–0.51), whereas the rate of miscarriage (14.6% vs. 9.2%) did not significantly increase (RR = 1.53, 95%CI 0.91–2.55). Conclusions: In DCTA triplet pregnancies, MFR to singleton pregnancy was associated with a reduced preterm birth rate and not associated with an increased miscarriage rate. Given the fact that the MC pair is reduced only to lower the rate of preterm birth, appropriate counselling and justification are important. In the absence of randomized controlled trials, data from systematic reviews are the best available evidence for counseling on the different management options. Full article

Dynamic Wireless Power Transfer (DWPT) is emerging as critical smart city infrastructure for sustainable urban mobility, enabling electric vehicle charging while driving. However, DWPT introduces complex fault scenarios requiring intelligent monitoring. Existing fault diagnosis approaches for wireless power transfer systems face three key complexities: (1) they are limited to static charging with only 2–4 fault categories, failing to address the time-varying coupling dynamics and segmented coil handover transients inherent in dynamic charging; (2) they lack integration with the host distribution grid, ignoring grid-side disturbances that propagate to charging stations; and (3) they offer only reactive detection without predictive capability for incipient fault management. This paper presents a deep neural network (DNN)-based fault diagnosis framework utilizing multi-station sensor fusion for DWPT systems integrated with the IEEE 13-bus distribution network to address these limitations. The system monitors 36 sensor features across three charging stations, employing feature-level concatenation with station-specific normalization for multi-station fusion, achieving 97.85% classification accuracy across eight fault types. Unlike static charging, the framework explicitly models time-varying coupling dynamics due to vehicle motion, including segmented coil handover effects. A digital twin provides dual-horizon prediction: long-term forecasting (24–72 h) for incipient faults and real-time detection under 50 ms for critical protection, with fault probability outputs and ranked fault lists enabling actionable maintenance decisions. The DNN outperforms SVM (92.45%), Random Forest (94.82%), and LSTM (96.54%) with statistical significance ($p<0.001$), while maintaining model inference latency of 4.2 ms, suitable for edge deployment. Circuit-based analysis provides analytical justification for fault signatures, and practical parameter acquisition methods enable real-world implementation. Five case studies validate robustness across highway, urban, and grid disturbance scenarios with detection accuracies exceeding 95%. Full article

With the collapse of the post-Cold War bipolar world order, religious institutions regained their public role in the socialist and people’s republic states of Central, Eastern and Southeastern Europe. Religion not only regained its social influence, but also once again became a decisive factor in shaping national identity. During the disintegration of the Socialist Federal Republic of Yugoslavia, religion did not merely attempt to fill the ideological void left by the crisis of the socialist value system; it also actively contributed to the reconfiguration of national values, culture, identity and political discourse. This study examines the religious factors that contributed to the sacralisation of national identity; the consolidation of the ‘Us’, ‘Them’, and ‘Us versus Them’ narratives; and the justification of wartime violence during the disintegration of the Socialist Federal Republic of Yugoslavia (SFRY). In this context, ‘Us’ refers to the dominant religious/ethnic community of a given member republic, while ‘Them’ denotes the ethnic majority and their confessional affiliations living in other member republics. This mainly refers to the three largest religious/ethnic communities, Orthodox Serbs, Catholic Croats, and Bosnia and Herzegovina Muslims. The ‘Us versus Them’ confrontation escalated tensions and ultimately played a central role in the disintegration of the SFR of Yugoslavia. The study concludes that religion played a dual role: on the one hand, it supported the preservation of community identity and social cohesion; on the other hand, it fostered exclusion, the ethnicisation of loyalty, the political instrumentalisation of religion, and the legitimisation of war discourses on the other. Full article

Functional analysis plays a critical role in early-stage aircraft system design by defining system functions that guide downstream architectural development and verification. In practice, many design deficiencies originate not from incorrect physical realization but from incomplete or ambiguous functional definitions established at conceptual stages. This challenge is particularly pronounced in aircraft systems, where interaction- and physical-effect-induced functions tend to remain implicit and weakly justified. To address this issue, in this study, we conduct a case-study-based comparative evaluation of three functional analysis paradigms: design-theory-oriented functional decomposition, systems-engineering-based functional allocation, and scenario-driven functional analysis. Using an aircraft ground deceleration scenario as a controlled context, this comparison examines how different function-derivation mechanisms influence the identification and justification of interaction- and effect-induced functions. Through structured cross-paradigm comparison, three distinct and, in principle, reproducible derivation mechanisms, namely decomposition-driven, responsibility-driven, and physical-effect-driven, are identified. In this study, the physical-effect-driven mechanism is examined through an effect-strengthened implementation of the scenario-driven paradigm. While all paradigms consistently identify mission-oriented functions, the examined scenario-driven implementation enhances transparency in functional justification and improves sensitivity to interaction- and effect-induced functions, thereby reducing the risk of omission during conceptual design. By formalizing these derivation mechanisms and clarifying their complementary roles, this study contributes to a clearer methodological understanding of functional identification in early-stage complex system design, while providing practical guidance for methodological selection and integration in aircraft system design. Full article

Background: Parents of preterm infants face significant emotional, informational, and caregiving challenges during neonatal intensive care unit (NICU) hospitalisation. Educational interventions are increasingly used to support parental readiness; however, considerable variation exists in their content, structure, and delivery. A clearer understanding of these components is essential to inform the development of effective, contextually responsive programmes. Aim: To identify and synthesise the core educational components, programme structures, and embedded parental support needs within NICU-based educational interventions for parents of preterm infants. Methods: A systematic search of peer-reviewed literature (January 2010–September 2022) identified 33 studies of high methodological quality. Data were extracted and synthesised using thematic analysis. Results: Three overarching domains were identified: (1) educational content, (2) programme structure and delivery, and (3) parental support needs integrated within educational delivery. The educational content encompassed the NICU environment, infant health and behaviour, caregiving practices, parental well-being, and discharge preparation. Programme structures varied widely in terms of intensity, duration, delivery modality, and facilitator roles, with limited justification for structural choices. Parental support–emotional, relational, and confidence-building–was inconsistently embedded despite evidence of its importance. Established interventions such as COPE, FICare, and FCC have clearer theoretical foundations and more holistic support than most locally developed programmes. Conclusions: NICU educational interventions positively influence parental knowledge, confidence, and parent–infant interaction; however, substantial variation and limited conceptual grounding hinder their comparability and scalability. The evidence base remains dominated by high-income settings, which limits its global applicability. Future research must prioritise theory-informed design, transparent reporting, and context-sensitive adaptation, particularly in under-resourced health systems, to support equitable and effective parental education for families of preterm infants worldwide. Full article