The present study examines the acquisition of English articles by Urdu L1 speakers in Pakistan, focusing on the use of definite articles. As seen in Hawkins and Towell (1994), the existing studies highlights the challenges that L2 learners face in case they do not have an equivalent article system in their L1, Master (1995) emphasizes the role of explicit instruction in facilitating article acquisition, and Ionin (2009) examines the role of the quality of input in L2 article learning, concluding that the amount and type of English exposure have a significant effect on how well learners master articles. However, it is found out that acquiring articles can be very difficult for L2 learners when their L1 is does not have an equivalent structure like the article-less languages of Urdu, Japanese and Korean. The current research explores whether the problem of article learning is similar in different age groups of participants i.e. adolescents (9-19 years old) and adults (20 years old or higher). For this purpose, 100 participants were taken. Following the Fluctuation Hypothesis (FH) by Ionin (2004), as a theoretical framework, the study identified the degree to which learners vary in their patterns of use of articles and how various input patterns influence proficiency. In addition, the research also examined the effect of level and grades of education at which start learning English, no of years of learning English and the effect of Mother tongue on the accuracy of the learners in the articles. The Data were collected by using elicitation test and a judgment task based on Ionin's (2004) methodology to assess learners' accuracy in using definite articles. The findings focus on how the age, educational level, and knowledge of languages interact and influence the acquisition of articles within L2 English regarding the stages when learners develop proficiency overcoming interference, overgeneralization and overuse. The findings not only enrich our understanding of the L2 article use, but also give practitioners, including L2 English teachers and learners, empirically based advisory on instructions design improvements and learning maximization.

Counting parts and wholes and the mastery of partitive language.

Bridging language models and knowledge graphs with controlled natural languages

Ontology-grounded knowledge graphs for mitigating hallucinations in large language models for clinical question answering.

Abstract Indigenous language knowledge can serve as a window into cultural identity and wellbeing. The aims of our study were twofold: (1) to capture the process of increasing awareness of Noongar to first-year speech pathology students between 2021 and 2024 at Curtin University, and (2) to develop Noongar Language and Speech Sound materials using feedback received from the Elders reference group, clinicians, educators, and speech pathology students. In doing so, we address the potential for misdiagnoses that currently exist when speech pathologists use standardised Australian English-based assessments to work with Noongar-speaking children in the areas of language, speech, and literacy difficulties. Drawing on Indigenous standpoint theory, we used Yarning through Kapati Time, Dabakarn-dabakarn , and Yindyamarra Methodology to build trust and relationships. Over four years, we co-designed language and speech sound materials through Participatory Action Research. The team included two Indigenous (Noongar, Wiradjuri) and three non-Indigenous academics. Two outdoor lectures and tutorials were delivered annually, reaching over 400 students, who contributed to the material design. Guided by the ethical framework by the Australian Institute of Aboriginal and Torres Strait Islander Studies, this Noongar-led project demonstrates how collaboration ensures cultural safety and promotes recognition of language differences during clinical assessment, reducing over- or underdiagnosis in multilingual Noongar-speaking children.

Causal discovery algorithms are often leveraged for inferring causal relationships and recovering a causal model from data. However, causal discovery from data alone is limited by the structural constraints of the used dataset, the lack of causal logic, and the lack of external knowledge. Thus, data-driven causal discovery can only suggest possible causal relationships at best. To overcome these limitations, Large Language Models (LLMs) and knowledge systems, such as Retrieval-Augmented Generation (RAG), have been proposed as alternatives to data-driven causal discovery and as a method to augment causal discovery algorithms. Using an expert-defined causal graph of chronic lower back pain, we further propose knowledge graph based RAG systems, such as GraphRAG, as an improvement over RAG systems for augmenting causal discovery (F1 0.745), benchmarking its performance against augmenting causal discovery with an LLM (F1 0.636), augmenting causal discovery with RAG (F1 0.714), and causal discovery alone (F1 0.396). We also explore the impact of different prompting methods for causality, such as querying for the plausibility of causal relationships, the presence of statistical associations, and the existence of temporal causal relationships, as inspired by the methodology of the domain experts constructing our ground truth. Lastly, we discuss how applications of LLMs, RAG, and graph-based RAG systems can impact and accelerate the causal modeling of chronic lower back pain by bridging the gap between domain knowledge and data driven approaches to causal modeling.

With the rapid development of artificial intelligence technology, ideological and political education in Chinese universities is undergoing profound change. This article is based on a three-layer concept of value resonance, situational perception, and adaptive presentation of educational resources. It explores how to use natural language processing, knowledge graphs, and reinforcement learning to construct an intelligent education framework to enhance the effectiveness of ideological and political education. This framework captures students' interest signals and emotional clues through multisource data, dynamically optimizes the order of resource presentation using a strategy network, and promotes value internalization and behavior transformation through a closed-loop feedback mechanism. The results show that the system demonstrated significant advantages at accelerating reinforcement learning convergence, accurately reaching emotional resonance, and enhancing classroom interaction. This study introduces new possibilities for the deployment of intelligent teaching systems.

In the study of human cognition, language is central in part because our species is the only surviving member of the primate lineage that is in possession of its distinctive properties. Evidence suggests that only in homo sapiens, spectacularly, did language emerge. Aside from the point of view of evolutionary origins, the other reason for language being a defining feature is its function as an indispensable instrument of higher order cognition. If language-related abilities are essential in this way, linguistics can be understood as a branch of psychology; and in general, the methods of cognitive science should be applicable to its study in every way. From another perspective, its centrality is related to the interconnectivity of language with the other higher order competencies of the mind. Thus, in all of the above, and in the following discussion, “language” should be understood as “knowledge of language.” As in all the branches of psychology, rooted in turn in biological science, attention to the relationship between applied research and theory cannot in the long run be neglected.

The purpose of the study is to analyse the results of the world’s first ‘Farinelli’ countertenor competition within the framework of the F. G. Handel International Festival, to outline the requirements for the vocal technique of modern countertenors, and promising ways of training Ukrainian singers to perform works of the countertenor repertoire in the musical content of world opera art. The research methodology is based on a comprehensive approach that includes the use of empirical approaches: observation and generalisation of the requirements for modern performers of baroque vocal music and modern repertoire, an analytical method used to highlight the history of countertenor singing and the results of the Farinelli countertenor competition, the importance of both baroque and modern repertoire in the process of educating a countertenor vocalist, his further creative activity. The methods used reveal the practical performing aspect of the research. The scientific novelty of the article lies in highlighting and summarizing the main requirements for the vocal technique of singing of modern countertenor performers, which affects the singer’s career growth and the content of his repertoire. For the first time in Ukrainian musicology, the first ‘Farinelli’ 2025 countertenor competition and the Ukrainian countertenor Vladyslav Shkarupilo, the youngest participant in the competition, are presented. Conclusions As a result of the research conducted the results of the first international competition of countertenors ‘Farinelli’ were analysed. The analysis of the works involved in the performance practice of the finalists of the competition makes it possible to identify a number of important requirements corresponding to the traditions of the performance of the musical aesthetics of the Baroque and the modern countertenor repertoire. Mastering the coloratura technique occupies one of the main places and should be considered not only as an element of the historical context of the performed works, but also as a necessary skill that reveals the technical capabilities of the voice and creates conditions for expanding the stylistic diversity of the repertoire. In addition to the rather complex vocal and technical training, an important role is played by the performer’s intellectual abilities, ability for analytical thinking, erudition and musical taste, knowledge of languages, musical styles and genres, general musical culture, and sociability. The Farinelli Vocal Competition is undoubtedly significant in world musical culture and has an impact on the development of performing vocal art. The competition provides an opportunity for young singers from all over the world, regardless of origin, education, or vocal schools, to demonstrate their talents to an international audience, and receive not only awards and engagements, but also well-deserved recognition from leading opera experts.

In recent years, Large Language Models (LLMs) have rapidly advanced in language understanding, reasoning, and generation, and are increasingly adopted as the “brain” of industrial intelligent systems. Nevertheless, in high‑risk and strongly regulated domains they still exhibit hallucination, weak domain grounding, limited interpretability, and privacy as well as security constraints. Knowledge graphs (KGs) encode domain entities, relations, rules, and events explicitly, providing controllable semantics and an explainable reasoning substrate. Retrieval‑augmented generation (RAG) injects external evidence into LLM prompting, while GraphRAG further introduces graph indexing and community‑level retrieval to preserve global structure and support multi‑hop reasoning. This review summarizes the evolution of LLMs, KG modeling and extraction, GraphRAG mechanisms, and a general fusion framework. Typical industrial applications are surveyed, and a coal mine flood emergency plan generation and evaluation approach is discussed to illustrate the practical value of graph‑grounded large models. KG‑enhanced retrieval also supports provenance tracking, allowing industrial users to audit the evidence behind model outputs.

In learning English as a foreign language, reading comprehension and grammatical knowledge are very important components, as they help students interpret information and construct meaning clearly. At the same time, they reinforce the capacity to produce the language. At the same time, they strengthen the ability to produce the language. In this sense, this quantitative correlational study examined the relationship between reading comprehension and grammar knowledge among students enrolled in a master’s program in EFL Pedagogy at a public university in Loja during the August–December 2025 period. For this purpose, standardized reading comprehension and grammar tests aligned with postgraduate academic requirements were administered. In this way, data was collected through online tests. The results showed a moderate positive correlation, indicating that students with higher grammar knowledge tended to achieve better reading comprehension scores. The highest performance was observed in inferential comprehension and sentence structures, while the lowest scores appeared in inferential comprehension and coherence. The main conclusion suggests that reading comprehension is closely linked to grammatical competence, reinforcing the theory. These findings highlight the need for targeted instructional strategies to strengthen advanced reading and grammar abilities among postgraduate EFL learners.

Background: Stroke is the fifth leading cause of death and the leading cause of long-term disability in the U.S. However, Deaf and hard-of-hearing (DHH) individuals face disproportionately low levels of stroke-related health literacy and encounter significant barriers to timely, effective care. Rochester, NY, has the highest population of DHH adults under age 65 in the country. According to the American Community Survey, about 40,000 people—roughly 3.5% of the local population—identify as Deaf or hard of hearing. A 2020 Journal of the American Heart Association study found over 60% of Deaf respondents couldn’t name a single stroke symptom, and only 61% said they would call 911 during a stroke or heart attack. Methods: This community-centered initiative is grounded in clinical stroke knowledge and academic training in American Sign Language (ASL) and Deaf Studies. The educational presentation was developed to explore barriers to stroke care in the DHH community and provide clinicians with practical tools to enhance communication and emergency response. Content includes basic ASL signs for emergency use, culturally competent communication strategies, and an analysis of systemic inequities affecting stroke care for DHH individuals. Results: Key barriers identified include: Inconsistent NIH Stroke Scale scoring for nonverbal or ASL-using patients, Communication challenges during emergencies-Worsened by interpreter delays, Limited provider awareness of Deaf culture and ASL-specific brain processing, Ineffectiveness of video interpreting in high-acuity settings, Insurance-related access issues Conclusion: By fostering inclusive and accessible stroke education, this initiative seeks to reduce health disparities, improve emergency responsiveness, and ultimately save lives. This presentation serves as a call to action for healthcare providers and educators to prioritize health equity by delivering culturally and linguistically appropriate education to underserved populations—including the Deaf and hard-of-hearing community. Generative AI (ChatGPT, OpenAI) was used to support language refinement. All content was reviewed and finalized by the author.

The article considers the theoretical and practical aspects of the formation of foreign language competence of students while reading. It is found that reading involves working with authentic texts of different subjects, complexity and promotes the development of foreign language communicative competence, forms readiness to speak a foreign language in a particular situation, and positively adapts to the further study of foreign languages. The factors that influence the successful development of reading skills in foreign languages are analyzed. The requirements for the selection of texts for processing are outlined and the main functions and tasks of the educational process that the teacher accomplishes in the process of reading foreign languages are determined. Traditional stages of working with foreign texts are described and a detailed description of each stage is provided. It is found that the assignments of the first pre-textual stage are aimed at modeling the background knowledge, which is necessary and sufficient for the reception of the proposed text, eliminating the semantic and linguistic difficulties of its understanding, and at the same time building of reading skills and abilities. The text tasks contain communication settings with specific instructions on the type of reading, the speed and the necessity of solving certain educational and communicative tasks. Exercises offered at the post-textual stage are aimed at testing the reading comprehension in order to control the degree of reading skills and the abilities to use the information obtained in future professional occupation. An algorithm and examples of assignments that can be used while reading at different stages of working with text material are offered. It is concluded that the proposed algorithm is not compulsory and universal. The teacher varies the assignments and the scheme of work with the text depending on the educational assignments, the topic and the level of knowledge of the foreign language by students.

With the widespread adoption of artificial intelligence, professionals in the design field are facing unprecedented transformation. On one hand, AI integration into design can significantly enhance workflow efficiency. On the other hand, AI has profoundly altered the design process, signifying a critical shift in human-machine collaboration. However, no research has yet explored how these changes in the design process impact the final design outcomes, nor how designers leverage their professional expertise to engage with AI during collaborative design processes. This study aims to clarify the respective roles of designers and AI by proposing an AI-enabled product design (AIPD) paradigm grounded in designers’ tacit knowledge. We divide AIPD into four phases: design preparation, large language model training, collaborative design, and knowledge base establishment. Subsequently, the research vessel design was developed based on this paradigm, with the quality of the design evaluated using the fuzzy comprehensive analysis method. The results demonstrate that this human-AI collaborative design process is effective in addressing product form, function, and structural challenges. This research elucidates the direction for human-machine collaboration in the AI context and proposes new pathways for AI application in product design.

Sign language recognition devices are gaining tremendous attention in recent years for helping speech and hearing-impaired community. The idea of fusing technology and sign language knowledge together to create a smart system is still being tried and developed all over the world with implementation with many different sign languages. In this paper, Malaysian Sign Language is given importance with 5 Malaysian Sign Language words being selected for recognition and prediction with new combination of sensor used compared to previous researches done for Malaysian Sign Language recognition and prediction. The combination of sensors used are 1 MPU9250, 1 MyoWare and 2 Force Sensitive Resistor sensors. 1D CNN time-series model is implemented with prediction accuracy ranging from 80 to 91 percentage.

The use of unggah-ungguh basa Jawa (Javanese speech levels) is a crucial aspect of Javanese language learning because it is closely related to social norms, politeness, and Javanese cultural identity. However, observations conducted at SMP IT Bina Amal Semarang indicate that many students still experience difficulties in applying unggah-ungguh basa Jawa, particularly in selecting appropriate vocabulary and using it in everyday communication contexts. This study aims to describe the difficulties faced by students in applying Javanese speech levels based on four main aspects, namely language knowledge, speaking ability, affective factors, and the learning environment. By employed a descriptive quantitative research design, this study involving 30 selected students. Data were collected using a questionnaire consisting of 20 statements and were analysed using a descriptive quantitative approach through the calculation of mean scores and percentages. The results reveal that the level of difficulty experienced by students in the aspect of language knowledge reached 64.55%, speaking ability 74.15%, affective aspects 66.00%, and the learning environment showed a relatively high level of difficulty at 81.80%. These findings indicate insufficient environmental support for the use of krama (polite Javanese) at home and in social interactions. The study concludes that students require more interactive, contextual, and applicative learning approaches to enable them to use unggah-ungguh basa Jawa more confidently and appropriately in various communicative situations

With the rapid advancement of generative artificial intelligence (AI) technology, college English writing teaching has ushered in new development opportunities. This study quantifies teaching effectiveness from two dimensions—external language knowledge mastery and internal writing thinking abilities, and conducts an in-depth exploration of the application value and practical effects of the generative AI technology ChatGPT in college English writing instruction. The research results indicate that ChatGPT can assume partial functions such as tool support and feedback evaluation across various stages of writing, significantly reducing teachers’ instructional burden. Additionally, the intelligent feedback system demonstrates remarkable advantages in improving the accuracy of students’ language expression and optimizing the logicality of text structure. Meanwhile, the application of ChatGPT is accompanied by inherent limitations, including lack of contextual awareness, potential for misinformation, and plagiarism risks, which necessitate teachers to play an irreplaceable role in evaluation and screening, rule supervision, and competency training. This study provides empirical references and targeted optimization suggestions for the practical path of AI empowering college English writing teaching.

Around 90-95% of deaf and hard-of-hearing (DHH) children are born to hearing parents, most of whom have no prior knowledge of a sign language. This creates a risk of language deprivation, with long-term effects on language, cognitive, and social-emotional development. Research shows that sign language input supports children's development without hindering spoken language acquisition. Yet many hearing parents of DHH children are discouraged from learning sign language by unfounded claims from professionals or by restrictive support systems. In the Netherlands, parental sign language courses are covered by health insurance, which reduces financial barriers but also reinforces a medical framing of deafness and sign language and limits parental choice by assigning families to a single provider. Higher-level sign language courses are scarce, and most materials focus on vocabulary rather than sentence formation, leaving parents at a basic level while their children progress rapidly. This brief highlights what is known about hearing parents' experiences learning sign language and calls for systemic changes: unbiased professional guidance, inclusion of DHH role models, greater parental choice in courses, and a socio-cultural view on sign language.

ABSTRACT In the context of digital education, this study proposes a data-mining–driven approach to improve key knowledge analysis in English teaching. A small-sample relational inductive reasoning model is developed by integrating graph neural networks (GNNs) with meta-learning to address data scarcity and personalised learning demands. An English teaching knowledge graph is constructed using digital textbooks, lesson plans, and authoritative language knowledge bases, with entity and relation extraction techniques used to model semantic relationships. To enhance inductive reasoning under limited data conditions, path masking and loss-weight meta-learning strategies are introduced. Experimental results show that the proposed model imports tens of millions of nodes in 38 seconds, achieving an 82% efficiency improvement over the basic GNN model. The multi-hop query delay is reduced to 6 ms, 98% lower than that of traditional databases, with significantly optimised memory usage. In practical teaching applications, students’ average reading scores increase by 8.2 points, grammar mastery reaches 92%, personalised learning path matching achieves 88%, and the repetition rate of incorrect answers decreases by 72%. Overall, the proposed method substantially enhances English teaching effectiveness and learning outcomes.

Many East or Southeast Asian toponymic doublets (including nesonymic units), as well as their graphic variants, are present in the Russian text vocabulary of the 18th-20th. due to different traditions of incorporating written exoticisms, though they do not always correspond to real geographical objects. This research focuses on Russian texts that include Taiwanese nesonyms. The primary objective is to identify the complete range of these units and, if feasible, to clarify their association with real or imaginary islands to ensure adequate reader reception or interpretation.To find the reasons for the use of doublets or variants in texts is also important. In the Russian vocabulary, the system of Taiwanese (by origin) historical and modern nesonyms includes hundreds of units and reacts to constantly updated geographical data or cultural and geopolitical events. On the other hand, this system preserves evidence of the ways in which Russians assimilated exoticisms in their texts. Moreover, toponymic variants and doublets in general, and nesonyms in particular, are not associated solely with the instability of the graphic, phonetic, lexical, or morphological norms. A pedantry in writing chains of doublets or variants can be explained by the conscientiousness of the translator, lexicographer, cataloger, navigator, and others, who desire to provide their readers with comprehensive information about the names of little-­ known topographic objects. The coexistence of nesonymic doublets and variants (that at first glance seems to hinder the establishment of literary or conventional norms for the use of exoticisms) signals the level of education of the author or translator and their knowledge of foreign languages. In a stylistic sense, the use of lexical doublets and graphic variants turns into a sign of educational texts.