Search Results (283)

This paper studies geometric scene formalization in vision-based educational sensing environments, where textual conditions and geometric diagram images jointly constitute heterogeneous perceptual inputs. The goal is to convert multimodal sensed information into standardized formal representations for machine understandable educational analysis. Existing methods remain limited by unstable cross modal alignment, inadequate expression of geometric relational constraints, and insufficient verifiability of generated outputs. To overcome these challenges, a unified modeling framework is proposed based on multimodal large models with structure-aware prompting and verification feedback. A geometry-oriented structure prompt injection mechanism is first introduced to encode prior cues of geometric entities, relational patterns, and constraint dependencies, which enhances the intrinsic alignment among textual descriptions, visually sensed diagram regions, and formal symbolic representations. In addition, an external verification feedback strategy is employed to constrain and iteratively refine the initial outputs, thereby improving structural consistency, syntactic correctness, and target proposition accuracy. To support this task, a new vision-based multimodal geometry formalization dataset is further constructed for model training and evaluation. Extensive experiments show that the proposed method can more effectively accomplish the transformation from multimodal sensed educational inputs to executable formal expressions, while also demonstrating stronger robustness and reliability in complex visual conditions. These results indicate that the proposed framework offers a feasible solution for structured scene interpretation, automatic problem analysis, error diagnosis, and intelligent feedback in vision-based educational systems. Full article

Previous research on Mandarin Chinese tones and intonation has focused primarily on universal sentence pitch patterns (declination) and sentence types (declarative and interrogative). The specific impact of internal syntactic structures remains under-explored. This study presents two acoustic experiments using controlled Tone 1 (high-level) stimuli to isolate intonational “big waves” from lexical “small ripples”. Experiment 1 investigates how syntactic position (subject vs. object), relative clause type (subject-relative vs. object-relative), and word class (verb vs. noun) influence pitch contours. Experiment 2 resolves conflicting findings regarding word-class pitch by testing nouns and verbs across four sentential contexts. The results indicate that subject positions carry significantly higher pitch than object positions, reflecting an interaction between SVO word order and declination. Crucially, subject-relative (SR) clauses exhibit a falling pitch tendency, while object-relative (OR) clauses show a rising trend. These results suggest that pitch realization is a complex “algebraic sum” of universal phonological trends, syntactic hierarchy, and semantic information structure. Full article

Informal urban settlements grow through incremental and adaptive processes, yet the temporal logic through which their access networks emerge, endure, and consolidate has received relatively little systematic attention. This paper examines the configurational development of the access network in Likoni, Mombasa, where rapid informal urbanisation has transformed an area containing only sparse footpaths into a dense urban network over two decades. Using historical satellite imagery, the study mapped five temporal states of access network for 2006, 2011, 2016, 2021, and 2026. The study utilises Space Syntax angular segment analysis. The analysis combines measures of angular connectivity, segment length, global and local integration, global and local choice, intelligibility, and synergy. The study aims to address three main questions: whether early informal footpaths persisted as the structural basis of later development of access network, whether subsequent growth strengthened local or global accessibility, and whether densification improved the overall configurational accessibility and legibility of the system as a whole. The results indicate that a finer-grained and more locally integrated network was produced through subdivision, densification, and the multiplication of short connecting segments. However, the gains were uneven across scales. Global integration and choice remained concentrated along a limited set of inherited and edge-related corridors, while local integration and local choice spread more widely through the settlement. The paper argues that the development of Likoni is a process of selective consolidation. Early footpaths became a persistent movement skeleton, forming the subsequent major paths of the later stages of the settlement. Later growth intensified local accessibility—albeit, as demonstrated through Space Syntax analysis rather than direct observation of movement—without necessarily producing notable improvements in global integration or whole-system configurational intelligibility. This finding adds a temporal and syntactic dimension to the understanding of informal morphogenesis. Full article

Large language models (LLMs) are increasingly being used in education, yet their correctness alone does not capture the quality, reliability, or pedagogical validity of their problem-solving behavior, especially in mathematics, where multi-step logic, symbolic reasoning, and conceptual clarity are critical. Conventional evaluation methods largely focus on final answer accuracy and overlook the reasoning process. To address this gap, we introduce a novel interpretability framework for analyzing LLM-generated solutions using undergraduate calculus problems as a representative domain. Our approach combines reasoning flow extraction and decomposing solutions into semantically labeled operations and concepts with prompt ablation analysis to assess input salience and output stability. Using structured metrics such as reasoning complexity, phrase sensitivity, and robustness, we evaluated the model behavior on real Calculus I–III university exams and compared it with the performances of students enrolled in the courses. Our findings revealed that LLMs often produce syntactically fluent yet conceptually flawed solutions with reasoning patterns sensitive to prompt phrasing and input variation. This framework enables a fine-grained diagnosis of reasoning failures, supports curriculum alignment, and informs the design of interpretable AI-assisted feedback tools. The framework was evaluated on Gemma 3, an open-access large language model, across zero-shot, retrieval-augmented generation, and contextual retrieval configurations, using nine real undergraduate calculus examinations from three course levels. To our knowledge, this is the first paper to apply a combined reasoning flow decomposition and prompt ablation framework to real undergraduate calculus examinations, benchmarked against actual student cohort performance, laying the foundation for the transparent and responsible deployment of AI in STEM learning environments. Full article

Classical information theory (CIT) characterizes the transmission limit for communication systems under syntactic accuracy, whereas semantic information theory (SIT) studies communication from the perspective of semantic fidelity induced by synonymous mapping. In this paper, we investigate the semantic channel capacity of Rayleigh fading channels under synonymous mapping of the channel gain and additive noise. We first derive the semantic capacity formula when synonymous mapping is applied to the channel fading coefficient and establish corresponding upper and lower bounds using Jensen’s inequality. To determine an optimized synonymous partition, the partition design is formulated as a constrained optimization problem and solved numerically using a neural network-based approach with the Adam optimizer. Furthermore, we extend the framework by applying synonymous mapping to both the channel fading coefficient and the additive noise and derive the corresponding semantic capacity formula together with its theoretical bounds. The numerical results illustrate the theoretical semantic channel capacity under synonymous mapping and validate the compatibility of the proposed framework with both CIT and SIT. At a 20-dB SNR with $K=8$ channel gain intervals and $J=4$ noise intervals, the semantic capacity reached 9.86 sebits/s/Hz. Full article

Modeling historically situated gender ideology remains challenging for language models, as contemporary embeddings struggle to reflect temporally specific semantic structures beyond surface lexical patterns. Although large language models exhibit extensive general-purpose performance, their direct use with history-specific semantic analysis is limited by the distributional mismatch between contemporary training data and historical linguistic patterns. These constraints encourage the distillation of temporally based semantic knowledge into small student architectures. To solve this issue, we propose Temporally Informed Distillation of Embedding Semantics (TIDES), which integrates continued pretraining on temporally specific corpora with feature-level distillation from large embedding teachers. We evaluate TIDES across teacher architectures with distinct pretraining objectives. While continued pretraining provides lexical and syntactic adaptation, our results show that improvements in ideological modeling cannot be attributed to additional training exposure alone. Rather, teacher–student structural alignment is also critical to transfer effectiveness. Contrastive, encoder-aligned teachers yield substantially more stable preservation of fine-grained, historically situated semantic distinctions. These findings suggest that temporal ideology transfer is representation-dependent: ideological meaning can be shaped by the geometry and training objectives of embedding spaces. By introducing TIDES and providing evidence that architectural compatibility can influence ideological inheritance, this study advances a representation-centered account of modeling ideology in temporally grounded semantic research. Full article

This study presents a newly developed Sentence Repetition Task/SRT as a tool designed to investigate the acquisition of different syntactic structures in children with typical development (TD) and Developmental Language Disorder (DLD). The tool is grounded in the Growing Trees (GT, henceforth) approach, which assumes that developmental progression reflects the hierarchical growth of the syntactic tree, as described in cartographic analyses of clause structure. The SRT Protocol was constructed following the three developmental stages identified by GT: VP/TP, lower zone of the Left Periphery (LP henceforth), and higher LP zone. A preliminary pilot version was administered to 27 TD and 28 DLD children, followed by a revised second version with improved item design and broader syntactic coverage, administered to 28 TD and 21 DLD children. Descriptive and inferential analyses demonstrate a clear hierarchy in the acquisition of Italian morphosyntax, fully consistent with the three-stage developmental progression predicted by the model. Children with DLD follow the same path but with delayed acquisition and slower consolidation of certain structures. These findings provide developmentally grounded benchmarks for identifying morphosyntactic delays and show that the SRT Protocol is a reliable tool for profiling early syntactic development. Crucially, the protocol supports diagnosis and clinical practice by helping clinicians ensuring interventions that are both theoretically informed and aligned with syntactic growth. Full article

As large language models (LLMs) scale to cover more languages, their potential to support low-resource settings becomes increasingly promising. However, the mechanisms underlying cross-lingual transfer and the factors that facilitate it remain insufficiently understood. Prior work has highlighted the role of linguistic similarity—particularly syntactic structure—in enabling transfer across languages. In this study, we present a broad empirical analysis of how multilingual LLMs encode and relate structural information across languages with varying typological properties. We combine multiple complementary methods, including hidden-state similarity analysis, typological correlation, probing for syntactic features, and attention-based structural comparisons, across four multilingual models and thirteen languages. Our findings show consistent correlations between representational similarity and syntactic relatedness, suggesting that structural properties of language influence how information is organized and shared across languages. We further observe that attention-derived structures exhibit partial alignment with gold-standard syntax, though this alignment should be interpreted as heuristic rather than direct evidence of syntactic encoding. Overall, our results provide a comparative empirical perspective on cross-lingual structural bias in multilingual LLMs and highlight the importance of careful methodological interpretation when linking representation geometry to linguistic structure. Full article

Classical Algorithmic Information Theory (AIT) provides a rigorous foundation for information-based similarity measurement, but classical formulations and their compression-based approximations largely operate at the syntactic level, making them sensitive to surface-level variation and insufficient for semantic equivalence. To address this limitation, this paper introduces Semantic Algorithmic Information Theory. The contributions are organized around three core aspects. First, regarding algorithmic extension, we formalize the Semantic Turing Machine System (STMS) to decouple abstract concepts from their diverse syntactic realizations. Within this framework, Semantic Complexity is defined as the minimum program length required to generate some realization in a synonymous set, thereby characterizing compact meaning representation. Second, to enable approximate computation, we move from the ideal, uncomputable semantic information distance to a model-based direct estimator of the Normalized Semantic Information Distance (NSID), which uses neural autoregressive models as conditional probability estimators. Finally, through experimental validation and comparative analysis, we show that the NSID estimator suppresses syntactic variance while preserving semantic structure. Empirical results indicate that NSID provides a practical, computable surrogate for semantic distance and improves upon classical syntactic metrics in evaluating cross-representational equivalence. Full article

Multimodal models such as CLIP and ALBEF essentially maximize cross-modal mutual information to align heterogeneous modalities, utilizing semantic consistency as an implicit prior. However, this alignment mechanism creates a structural vulnerability: the models rely heavily on invariant information coupling. In this work, we investigate this vulnerability and propose a symmetry-driven adversarial attack framework. Unlike standard methods that inject high-entropy unstructured noise, our approach designs collaborative perturbations by modeling semantic-consistent mappings between geometric image transformations and syntactic text variations. By explicitly exploiting the information redundancy inherent in cross-modal symmetries, our method effectively reduces the entropy of the adversarial search space. This reveals a fundamental trade-off between information invariance and robustness, achieving state-of-the-art attack success rates with imperceptible perturbations. Full article

Large language models (LLMs) face difficulties in leveraging the syntactic structures and entity relations embedded in text for long-document information extraction. To address this issue, this paper proposes a generative extraction method integrating heterogeneous topology awareness and spatial alignment. The method first extracts syntactic and coreference information to construct a heterogeneous document graph and employs a mixture-of-experts network to decouple and encode multi-type topological features. A component orthogonal projection mechanism and a graph-text contrastive learning strategy are then utilized to align the extracted graph features to the underlying semantic space of the language model with high fidelity. Furthermore, Topology-Aware Encoder compresses the global features into fixed-length structural prompts to guide text generation. Experiments on the ACE2005, WikiEvents, and DuEE datasets demonstrated that the proposed method achieved state-of-the-art performance on information extraction tasks. Consequently, these results suggest that the proposed framework is a promising approach for complex information extraction across base LLMs of different scales. Full article

Voice cloning enables realistic fake speech in which a speaker’s identity is preserved while the spoken message is semantically altered. This paper asks whether such meaning-level manipulation leaves detectable traces in transcripts alone. To study this problem, we introduce FakeSpeech+, a paired real–fake dataset built from authentic speech clips and their matched semantically altered counterparts, re-embedded into cloned voices while preserving speaker identity. Using this dataset, we conduct a transcript-first analysis based on interpretable text-only features from two groups: (i) linguistic content organization and discourse dynamics, and (ii) compact production-related proxy cues, including hesitation and disfluency markers. We evaluate these cues under transcript-length control through residualization and compare authentic and manipulated transcripts using statistical and experimental analyses. The results show that only a limited subset of features retains strong separation after length control, with coordination-related structure and emotion anchoring emerging as the clearest cues, while several production-related and discourse-variability features show weaker but still informative differences. In contrast, a number of syntactic, lexical-diversity, and other discourse-level features show substantial overlap after residualization. These findings indicate that transcript-level structure and selected production-related cues remain informative under realistic content-manipulation threats, supporting the value of transcript-based analysis for identity-preserving fake speech. Full article

Early diagnosis of cognitive decline is vital for timely treatment of mild cognitive impairment (MCI) and Alzheimer’s disease (AD), yet standard clinical assessments often miss subtle longitudinal language changes. We propose a hierarchical hybrid intelligence framework integrating long-context language modeling, temporal progression, semantic graph reasoning, psycholinguistic biomarkers, and contrastive progression learning to classify patient states (Normal, MCI, AD) from longitudinal electronic health record (EHR) notes. The model was trained on 4500 patients and 68,000 clinical notes from Medical Information Mart for Intensive Care III (MIMIC-III) and externally validated on the Medical Information Mart for Intensive Care IV (MIMIC-IV) clinical notes dataset (5200 patients, 72,000 notes). Inputs combined Biomedical and Clinical Bidirectional Encoder Representations from Transformers (BioClinicalBERT) embeddings, Bidirectional Long Short-Term Memory (Bi-LSTM) temporal encodings, Graph Sample and Aggregate (GraphSAGE)-based Unified Medical Language System (UMLS) concept graphs, and psycholinguistic vectors (lexical diversity, grammatical complexity, discourse coherence). On the MIMIC-III hold-out set, the model achieved 99.999% accuracy, a macro F1-score of 0.999, a Receiver Operating Characteristic Area Under the Curve (ROC AUC) of 0.999, and a temporal stability variance of 0.0008. Monte Carlo cross-validation (10,000 folds) yielded $99.997\pm 0.003$% accuracy and $0.999\pm 0.001$ macro F1. Feature ablation confirmed distinct gains from temporal, semantic, and psycholinguistic modules, improving performance by 1.1% over text-only baselines. Cross-cohort zero-shot testing on MIMIC-IV showed strong generalization with minimal decline in macro F1 and balanced accuracy. Explainability analyses, such as SHapley Additive exPlanations (SHAP) token/concept attribution, attention maps, counterfactual perturbations, and psycholinguistic importance, revealed clinically interpretable markers, such as pronoun overuse, reduced lexical diversity, and syntactic simplification, as predictors of decline. Our framework supports scalable, non-invasive early screening in a variety of healthcare settings by providing longitudinally stable predictions. Full article

The current literature on electrical system maintenance highlights three technology domains—Building Information Modeling (BIM), Digital Twin (DT), and extended reality (XR)—that have independently demonstrated strong potential for improving lifecycle information management, predictive analytics, and operational support. However, their convergence remains largely underexplored, particularly in electrical system maintenance. This paper provides a structured review of BIM–DT–XR convergence in electrical system lifecycle management, examining their roles across lifecycle phases and their integration through literature synthesis and cross-domain implementation evidence. BIM is analyzed as a basis for modeling and integrating facility management with electrical asset lifecycles; DT as a framework for dynamic system representation and applications in electrical and power systems; and XR as a means of visualizing and interacting with BIM-DT environments. Cross-domain implementation evidence from an industrial electrical facility and a tertiary smart-building pilot shows that BIM–DT–XR integration is technically feasible at pilot scale. However, the analysis identifies five structural integration gaps: semantic misalignment between building-oriented IFC and grid-oriented CIM ontologies; fragmented standard adoption; inconsistent data governance and naming practices; validation approaches focused on syntactic rather than dynamic model fidelity; and the separation of XR visualization from predictive DT capabilities. The implementation evidence further indicates that real-world deployment remains constrained by data quality limitations, integration complexity, cost factors, and interoperability with legacy systems. The review concludes that, despite the maturity of individual technologies, their effective application depends on advances in semantic alignment, lifecycle data governance, validation of dynamic models, and scalable integration frameworks, enabling the transition toward integrated, interoperable, and lifecycle-aware infrastructures for electrical system maintenance. Full article

In research on verbal periphrases, analyzability constitutes a central parameter, both for describing the grammaticalization processes to which these constructions are subject and for defining their categorical status. This paper focuses on a specific verbal periphrasis: [tardar en + infinitive]. Its historical development is examined, along with the recent emergence of a dative of interest in this construction, drawing on quantitative data from various digital corpora. The findings show that over time en became the predominant linking element between the auxiliary and the infinitive and that the order of the components of the periphrasis gradually became fixed. The data also reveal that the new pattern with the dative of interest occurs more frequently in informal written language and colloquial registers, where the object pronoun contributes to clarifying the construction’s potentially opaque meaning. We argue that grammaticalization has reduced the syntactic analyzability of the construction, whereas the incorporation of the dative of interest points to speakers’ perception of tardar as an independent verb, thereby reflecting increased analyzability. This case study illustrates that the analyzability of a construction is not necessarily unidirectional, but may fluctuate over time, shifting in different directions at distinct historical stages. Full article