Search Results (2,322)

In the context of industrial digitalization and the widespread adoption of process automation systems, Knowledge Management Systems (KMS) play a key role in providing operational personnel with up-to-date instructions and regulations. However, the inherent ambiguity of natural language in technical documentation remains a serious obstacle, leading to incorrect operator actions, process deviations, and increased safety risks. This article investigates the integration of Large Language Models (LLMs) into KMS and its impact on user experience and human–machine interaction in industrial automation environments. A method called Semantic Latent Choice Detection is presented, designed to systematically identify interpretation ambiguities in process instructions and operator commands. Unlike existing approaches that require access to the internal model architecture (“white box”) or token-level logits, the proposed method is logit-free and operates with closed commercial LLMs (“black box”) via standard API interfaces. The method analyzes the semantic similarity of binary text blocks and polysemous terms within the context of a specific technological process. Using a metallurgical production case study, we demonstrate how the system detects hidden semantic collisions (e.g., the difference between “adding ferroalloys into the ladle” and “feeding ferroalloys onto the conveyor”) that are missed by traditional rule-based validation methods. Instead of arbitrarily selecting an interpretation, the system initiates a clarification request to the human operator, thereby reducing cognitive load, preventing erroneous automated decisions, and increasing trust in the KMS. An empirical evaluation conducted in a real-world industrial setting (unit control rooms and dispatch centers) shows a statistically significant reduction in errors related to misinterpretation of process regulations. The article contributes to the fields of automation engineering, knowledge management, and human-centered automation by proposing a novel method for validating operational instructions in high-risk industrial environments. Full article

Human placenta-derived amniotic membrane (hAM) and its derivatives have attracted growing interest as bioactive materials for tissue engineering, regenerative medicine, and biomaterial design. This review summarizes the anatomical and cellular characteristics of hAM and examines the interplay between its regenerative and immunomodulatory properties. Key hAM-derived biomolecules are discussed, with an emphasis on their roles in immune regulation, angiogenesis, extracellular matrix remodeling, and tissue repair across diverse regenerative contexts. Current applications of hAM-based materials in tissue engineering and regenerative biomaterials are reviewed, including emerging studies involving soft tissue applications. In addition, recent efforts to integrate hAM derivatives into 3D bioprinting approaches are examined, including their use as bioink components or biofunctional additives to hydrogel-based systems. Although studies involving hAM-based bioprinted constructs remain limited, the available findings suggest promising regenerative, proangiogenic, and bioactive effects. Challenges related to material processing, printability, standardization, and reproducibility are also discussed. Overall, this review highlights the biomaterial potential of hAM derivatives and their emerging relevance to biofabrication strategies aimed at developing biologically instructive constructs for regenerative medicine. Full article

Sustainability literacy encompassing environmental knowledge, systems thinking, and communication competence remains underdeveloped in higher education despite its critical importance for environmental advocacy. This sequential explanatory mixed-methods study examined associations between technology-mediated genre writing development, integrated with infographic design, and changes in sustainability literacy. Drawing on sociocultural activity theory and genre-based writing pedagogy, the study examined 320 students across 3 timepoints using structural equation modeling and qualitative interviews. Findings reflect an association and a statistically significant growth in Sustainability Literacy (2.91 to 3.54, p < 0.001), with Communication Competence showing the largest gain. Structural equation modeling reflects that Technology Integration was associated with Sustainability Literacy both directly and indirectly through Genre Writing Competence. Qualitative analysis identified three mechanisms: digital tools visualizing sustainability systems, genre awareness developing through audience-focused design, and communication efficacy fostering environmental values. The study’s novelty lies in showcasing that when students integrated infographic design with genre writing instruction, there was reported growth in students’ self-rated multidimensional sustainability literacy experience; not merely knowledge accumulation but also communication competence and systems thinking requisite for authentic environmental action. Full article

Although GenAI has been increasingly applied in pre-service teacher education, limited evidence is available on how permitted but unscaffolded GenAI use affects pre-service teachers’ learning and professional development in authentic course contexts. Grounded in cognitive load theory and the zone of proximal development, this quasi-experimental study examined the effects of unscaffolded GenAI use in an 11-week instructional design course. Two intact sophomore classes at a normal university participated, with one class permitted to use GenAI without prompt templates or instructional guidance and the other not permitted to use GenAI. Data were analyzed using paired-samples t-tests and a one-way analysis of covariance (ANCOVA). After controlling for pretest scores, no significant group differences were found in AI readiness, self-regulated learning, or critical thinking, whereas the control group showed stronger instructional design performance. Within-group comparisons showed that both groups improved in AI readiness and instructional design performance, but not in self-regulated learning or critical thinking. These findings suggest that, in this course context, unscaffolded GenAI access alone may be insufficient to support pre-service teachers’ professional learning and may be less favorable for their instructional design performance. Full article

The rapid integration of Generative Artificial Intelligence (GenAI) into programming education has raised important questions regarding its impact on learning processes, conceptual understanding, and technological dependency. This study analyzed the effects of four GenAI-supported instructional strategies in an introductory programming course for undergraduate engineering students. A multi-group quasi-experimental pre-test–post-test design was implemented involving 686 students distributed across 53 class groups, from 10 campuses, taught by 32 professors. The instructional conditions included Quizzes for Self-Regulation, Github-Copilot-assisted learning, Prompt Problems with Iterative Refinement, and Flipped Learning enhanced with GenAI, which were compared against a traditional teaching approach. Learning outcomes were measured using normalized learning gain, while statistical analyses were conducted using non-parametric methods due to deviations from normality and heteroscedasticity. Results indicate that GenAI integration did not produce statistically significant overall differences in learning gain when all GenAI-supported strategies were analyzed as a single cluster compared to traditional instruction. However, differences emerged between specific strategies, with Quizzes and Copilot-based approaches having higher median learning gains than Prompt Problems and Flipped Learning strategies. No statistically significant differences associated with gender were identified. These findings suggest that the effectiveness of GenAI in programming education depends less on the mere presence of the technology and more on the pedagogical conditions under which it is integrated into the teaching–learning process. Full article

Artificial intelligence is rapidly changing education. However, many learning chatbots are still reactive tools, which respond to arbitrary questions without leading learners through a meaningful pedagogical journey. This article presents a deterministic state-machine orchestrator coupled with a local large language model and a knowledge-graph-framed tutoring strategy for personalized education. The proposed virtual tutoring agent is designed to combine the flexibility of conversational AI with the reliability of explicit instructional states, key performance indicator (KPI) tracking, learner profiling, and controlled transitions between explanation, practice, feedback, assessment, and remediation. The system is not meant to replace the teacher, but rather to act as a teaching co-pilot that provides ongoing feedback, personalized learning paths, accessibility, and safer deployment by processing data locally. The study also presents a compact interview-based evaluation framework and statistical analysis of user perceptions across interactivity, individuality, proactivity, security, accessibility, gamification, and global preference for educational agents over classical chatbots. The findings show that learners appreciate personalized and interactive support and that proactivity is the key feature that distinguishes an educational agent from a regular chatbot. With this article we argue that deterministic orchestration can help make AI tutoring more transparent, controllable, and ethically fit for real learning contexts. Finally, it discusses privacy, educational value, limitations and future improvements to be made before the large-scale adoption of such systems. Full article

Widening participation has expanded access to higher education, yet disparities in retention and completion among mature students persist, indicating limits in existing structural and engagement-based explanations. This study addresses this gap by developing Behavioural Alignment Theory (BAT), a framework that conceptualises persistence as an emergent outcome of alignment between motivational regulation, expectancy recalibration and instructional architecture. Using a conceptual integration and theoretical mapping methodology, the article synthesises Self-Determination Theory, Expectancy–Value Theory and Cognitive Load Theory, grounded in foundational and contemporary literature on mature and widening access students. The analysis shows that persistence is shaped through dynamic interaction between institutional design and behavioural processes. Evidence indicates that misalignment across these domains can destabilise engagement, while coherent alignment supports sustained participation under conditions of role complexity and constraint. The study concludes that persistence is not reducible to individual attributes or isolated institutional factors, but emerges from system-level interaction between psychological processes and institutional conditions. The contribution of BAT lies not in the invention of new constructs, but in providing a mechanism-explicit mid-range integrative framework that specifies how established constructs interact within institutional settings to shape persistence or withdrawal. Full article

Amid the rapid advancement of artificial intelligence (AI) and the digital transformation of schooling, computational thinking has become a foundational competency for K-12 learners and an organizing principle for AI education. Existing research suggests that students need not only programming knowledge, but also the ability to analyze problems, reason with data and models, and evaluate intelligent systems responsibly. However, current K-12 AI education often remains fragmented, with insufficient curriculum progression, tool-oriented instruction, uneven teacher preparation, and limited attention to learning processes, transfer, and ethical AI use. As a conceptual and integrative framework article, this paper synthesizes policy guidance and recent instructional research on integrating computational thinking into K-12 AI education. Building on prior Chinese scholarly discussion of Computational Thinking 2.0, we adopt and extend this perspective to connect rule-based algorithmic reasoning with data- and model-centered AI problem solving. We propose a teacher–student–AI collaborative framework and a multi-level pathway covering tiered curriculum design, blended human–AI teaching, scaffolded project-based learning, AI agent-supported feedback, teacher development, and process-oriented, transfer-sensitive, and ethics-aware assessment. We argue that AI education should move beyond tool demonstration and code production toward authentic problem solving, model understanding, responsible human–AI collaboration, and reflective innovation. The framework offers guidance for cultivating students’ computational thinking, AI literacy, and creative problem-solving capacity, while identifying directions for empirical validation. Full article

Architectural design studios increasingly emphasize sustainable, people-centered outdoor space, yet students often treat vegetation as ornament rather than as a spatial and environmental device, rarely translating spatial analysis into explicit design decisions. This study examines how intermediate-level architecture students translate space-syntax indicators—choice/betweenness, local integration and visual integration—into strategic vegetation decisions for paths, pause areas, visual filters and comfort in a minimal sustainable shelter. Using an exploratory mixed-methods design, fourteen anonymized student sheets from a design-studio examination at Universidad Técnica Particular de Loja (Ecuador), located in Pucará Park, were assessed with a five-criterion analytic rubric (scored out of 1.00), complemented by content analysis coding nine vegetation functions. The mean score was 0.84 (SD = 0.06), with high internal consistency (Cronbach’s α = 0.93). Achievement differed across criteria (Friedman test, p = 0.014): graphic clarity was highest (87.1%) and the reading of spatial analysis—especially operationalizing choice/betweenness—lowest (82.5%). Spatial-analysis and vegetation-function scores were positively associated (Spearman’s ρ = 0.61). Coupling space syntax with strategic vegetation offers a replicable, evidence-based pedagogical model, while indicating that operationalizing configurational indicators requires more explicit instructional scaffolding. Full article

This study investigates the effectiveness of peer assessment as a formative learning strategy in multilingual English-Medium Instruction (EMI) higher education. Although peer assessment is widely associated with enhanced learner engagement, self-regulation, and feedback literacy, its implementation in linguistically diverse EMI settings remains insufficiently explored. Addressing this gap, the study examines students’ perceptions of peer assessment, the degree of alignment between peer and instructor evaluations, and students’ reflective experiences in a multilingual university context. A mixed-methods design was employed with 60 undergraduate students enrolled in EMI courses at South East European University in North Macedonia. Quantitative data were collected through a structured questionnaire and peer-assessment rubric, while qualitative data were obtained from reflective interviews with 37 students. Quantitative data were analyzed using descriptive statistics, Pearson correlation analysis, and paired-samples t-tests, whereas qualitative responses were examined through thematic analysis. The findings reveal generally positive student perceptions of peer assessment, particularly regarding self-reflection, critical thinking, collaboration, and confidence in evaluating academic work. Significant positive correlations among key dimensions of feedback literacy suggest that peer assessment supports interconnected cognitive and metacognitive learning processes. Although a statistically significant difference emerged between peer and instructor scores, overall agreement was moderate, with students tending to assign slightly lower marks than instructors. Qualitative findings further indicate that peer assessment enhanced students’ understanding of assessment criteria and learning processes while also exposing challenges related to language proficiency, emotional discomfort, and concerns about feedback accuracy. The study demonstrates that peer assessment can be an effective pedagogical approach in multilingual EMI classrooms when supported by clear assessment criteria, structured guidance, and feedback training. These findings contribute to research on feedback literacy, formative assessment, and multilingual learning in higher education. Full article

Despite the recent surge in research on the use of AI in English language learning, little attention has been paid to its role in improving phonological awareness among preschoolers. Most existing studies have focused on general literacy skills or older learners, with insufficient emphasis on early phonemic awareness and its subskills. Furthermore, there is a lack of research examining these relationships within Arab or multilingual contexts. This study investigates the impact of artificial intelligence (AI) tools on the development of English phonological awareness in kindergarten children in an Arab educational context in Jordan using a quasi-experimental design. The participants comprised 45 students divided into two groups: a control group (n = 23), consisting of 14 females and 9 males, and an experimental group (n = 22), consisting of 12 females and 10 males. All participants were physically and mentally healthy 5–6 year-old children from similar socioeconomic and cultural backgrounds. The experimental group was taught via the AI-based Starfall platform and the control group was taught via conventional teacher-oriented instruction. Both groups were given pre- and posttests, which included assessments of five phonemic awareness skills: initial sound recognition, blending, segmentation, deletion, and substitution. Descriptive statistics, including means and standard deviations, and independent-samples t-tests were calculated to determine the effect of the AI program on developing kindergarteners’ phonemic awareness compared with conventional teaching methods. The findings of the study show significant improvements in the experimental group compared with the control group. Bringing AI into the kindergarten classroom may improve literacy instruction and, in turn, early reading readiness through engaging, interactive and adaptive learning experiences. Full article

The expansion of China’s modernized eco-environmental monitoring networks has generated vast amounts of data. Consequently, traditional, expertise-reliant analysis is increasingly ill-suited for agile regulatory decision-making. Although large language models (LLMs) present a promising alternative, their practical deployment remains limited by domain-specific knowledge gaps, hallucinations and an inherent difficulty in managing multi-faceted ecological tasks. This study introduces EnvSentry, a novel multi-engine collaborative LLM framework designed for intelligent eco-environmental monitoring and governance. EnvSentry coordinates reasoning, instruction, and multimodal engines, supported by a dynamic, vector-indexed knowledge base and retrieval-augmented generation (RAG) to ensure factual veracity. By transitioning operational workflows from fragmented, latent batch processing to integrated, real-time intelligent agent chains, the system achieves a closed-loop capability of intent recognition, data retrieval, and quality control. The model was evaluated across distinct environmental contexts, specifically water quality anomaly detection and air quality forecasting. Results show that EnvSentry yields higher analytical precision and attribution rates than baseline methods, while compressing decision-making latency from hours to seconds. Relative to baseline models, EnvSentry achieves a 25% improvement in water quality attribution accuracy (50% to 75%), a 90% reduction in decision making latency for anomaly detection, and a 10% absolute gain in data anomaly detection accuracy. In air quality forecasting, it reduces expert judgment time from 60 to 20 min and attains >85% agreement with expert forecasts when used by non-specialist personnel. These improvements suggest a practical shift in eco-environmental monitoring—moving from fragmented, reactive measures toward an integrated and proactive system. Consequently, this approach offers a viable path toward data-driven autonomous ecological management. Full article

This study explores the application of artificial intelligence (AI) in higher education to enable the early identification of at-risk students using only engagement data from learning management systems (LMS). Unlike many existing early-warning models that are limited to single disciplines or rely on sensitive demographic and prior academic records, the proposed approach offers a privacy-conscious and highly generalizable predictive framework suitable for diverse higher education contexts. The dataset includes over 1.7 million LMS interaction records from 236 undergraduate subjects spanning four academic divisions. These subjects encompass a wide variety of instructional designs and assessment structures. To address cross-subject heterogeneity, this study employs rank-based engagement features that represent students’ relative behavioral patterns within each course, facilitating meaningful comparison across disciplines without reliance on absolute activity levels. Using standard machine learning classifiers, the model achieves over 90% prediction accuracy for final subject performance by Week 3 of the semester, demonstrating that reliable early detection of at-risk students is feasible at an early stage of teaching and learning. Rather than claiming intervention effectiveness, the study positions AI-enabled early prediction as a scalable foundation for proactive student support and enhanced teaching responsiveness, with the potential to inform timely pedagogical actions such as targeted outreach and academic advising. By emphasizing generalizability, ethical data use, and privacy protection in AI-enabled learning analytics, this research contributes practical insights into how predictive AI can responsibly support teaching and learning in higher education. Full article

Although multimodal large language models (MLLMs) have achieved remarkable progress in visual question answering, they remain limited in tabular tasks that require fine-grained structured information perception and complex logical reasoning. This limitation primarily stems from the high density of structured information inherent in tables and the scarcity of high-quality instruction tuning data. To address these challenges and improve the model’s reasoning accuracy in tables, we propose MMTR, a strategy-guided multimodal table reasoning method with reflective self-correction. Mechanistically, we design a dual-LoRA architecture: the Strategy LoRA is responsible for generating structured reasoning steps, while the Reflection LoRA verifies and self-corrects these initial outputs. Their synergy empowers the model with a closed-loop capability of “reasoning–reflection–correction”. On the data front, we construct StrTab-QA, a large-scale dataset comprising question-answering, negative, and reflection samples, providing diverse supervision signals. During training, we further introduce a progressive “reasoning-to-reflection” fine-tuning strategy to gradually achieve cross-modal alignment and structural adaptation. Furthermore, coupled with an adaptive resizing and padding scheme, our approach effectively preserves table structures and minimizes information distortion during visual encoding. Extensive experiments demonstrate that MMTR consistently outperforms strong baselines across multiple table reasoning benchmarks. Full article

This study examines how agent type (customized vs. general-purpose) and feedback style (Socratic vs. directive) are associated with learners’ engagement with artificial intelligence (AI)-generated feedback in self-directed learning (SDL), with particular attention to patterns in feedback quality, self-regulatory behaviors, learning experiences, and learning outcomes. A 2 × 2 mixed factorial experiment was conducted with 51 postgraduate students who completed two instructional design tasks under different feedback conditions. Quantitative results indicated that customized agents generated feedback with higher accuracy and specificity than general-purpose agents. Socratic feedback was associated with stronger comprehension monitoring, whereas directive feedback was associated with higher cognitive load. A significant interaction suggested that the advantage of customized agents in learning outcomes, operationalized as short-term task improvement, emerged under directive feedback but not under Socratic feedback. Qualitative analysis indicated that Socratic prompts encouraged deeper, logic-oriented reflection, whereas directive feedback provided actionable guidance that facilitated task completion. Learners adopted feedback selectively based on perceived accuracy, and trust in customized agents was higher when feedback was clear and contextually aligned. These findings suggest that the effectiveness of AI-generated feedback is shaped not only by agent type and feedback style but also by how learners evaluate and use feedback. Full article