Search Results (457)

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

Offensive language detection is crucial for ensuring safe and inclusive digital environments. Identifying harmful content protects users and supports healthier online interactions. Despite advances in transformer-based models, particularly Large Language Models (LLMs), their application to this task remains underexplored for low-resource languages such as Moroccan Arabic, especially compared with high-resource languages. This study evaluates the performance of various open- and closed-source LLMs for offensive language detection in Moroccan Darija. The evaluated models include general-purpose LLMs such as LLaMA, Mistral, and Gemma, as well as Arabic-focused models such as ArabianGPT, Falcon Arabic, and Atlas-Chat. We also experiment with reasoning models such as DeepSeek and GPT-4. Beyond traditional evaluation metrics, we investigate the robustness of these LLMs and examine the impact of adversarial training on their performance. Moreover, we contribute to the field by creating a large, high-quality dataset. Our evaluation revealed that GPT-4o Mini achieved the best overall performance, reaching an F1-score of 88%. However, robustness testing under black-box and white-box adversarial attacks exposed notable vulnerabilities, with attack success rates reaching 30%, thereby highlighting the need for enhancement. Despite the complex morphology and linguistic variability of Moroccan Darija, adversarial training resulted in a notable improvement in both overall model performance and robustness against adversarial attacks, yielding an average increase of 20.89% in resistance to attacks. Furthermore, this approach enabled GPT-4o Mini to achieve an F1-score of 91%, surpassing the current state-of-the-art performance by 6%. These results highlight the importance of incorporating adversarial approaches in low-resource dialectal settings to effectively address linguistic variability and data scarcity. Full article

Background: Communication, language, and speech disorders are highly prevalent in children with cerebral palsy (CP) and substantially impact social, educational, and community participation. However, few studies have systematically characterized communicative and linguistic profiles using standardized assessments. This paper outlines the work of the ‘Italian CP & Language Network’ over the last two years, focusing on identifying research priorities, developing specialized assessment protocols, and proposing a shared classification system for speech and language disorders in children with CP. Methods: A survey was sent to 11 specialized centers to investigate clinical practices and assessment tools. Based on the results and an extensive literature review, the group developed three age- and complexity-based diagnostic protocols and a shared classification system. Results: The survey highlighted high variability in test selection, especially for speech and pragmatic assessment, and a significant need for ad hoc tools for augmentative and alternative communication (AAC). Three standardized protocols were defined: (1) early language (<48 months), (2) school-age language and pragmatics (4–12 years), and (3) minimally verbal children (6–12 years). A multi-level classification system for language and speech disorders was proposed to improve diagnostic consistency. Conclusions: Standardizing assessment is a critical step toward early identification of communicative vulnerabilities to guide tailored interventions and promote participation and quality of life across developmental stages. The group provides a framework for prospective multicenter data collection to correlate linguistic and speech phenotypes with neuroradiological features and motor outcomes. Full article

Quality evaluation in production systems represents a significant challenge in the manufacturing industry, particularly in environments where expert judgment plays a key role in managing the inherent uncertainty of the production system. This study proposes a fuzzy multicriteria decision-making control chart, termed Fuzzy Decision-Making Control Chart based on AHP-Extent and Triangular Fuzzy Numbers (FDMCC-AHPE). The method integrates expert knowledge through triangular fuzzy numbers and a Fuzzy Analytic Hierarchy Process supported by Extent Analysis, to define fuzzy decision intervals for quality assessment and subsequently perform a structured analysis to classify the product within a control chart framework. In this framework, expert judgments expressed through linguistic evaluations are systematically translated into triangular fuzzy numbers and processed using FAHP–Extent Analysis, allowing the aggregation of subjective assessments within a structured mathematical decision model. The proposed method was validated in a tannery company, specifically in the retanning process. The industrial case study considers both qualitative criteria, such as surface defects and color uniformity, and quantitative process variables that include bath pH, treatment duration, and processing temperature. The results were compared with an empirical expert-based evaluation and a structured expert assessment supported by a multicriteria decision-making method. The findings demonstrate that the FDMCC-AHPE exhibits greater sensitivity in discriminating between quality states under uncertain evaluation conditions, particularly when samples involve complex evaluation conditions. Full article

Background: Reading is a fundamental skill for children’s cognitive, social, and academic development which relies on the integration of multiple linguistic and cognitive abilities. Longitudinal studies consistently show that oral language skills predict reading development both in typically developing children and in those at risk for reading difficulties (RD). Despite strong empirical evidence, a gap remains between research and educational practice. Objective: The present study aims to compare linguistic variables, including vocabulary, oral text comprehension, oral morphological awareness (OMA), and morphological skills between diagnostic group (control vs. at-risk), study grade, and sex. Method: The study included 93 Spanish-speaking children aged 6 to 12 years (M = 8.7, SD = 1.9; 50 boys, 43 girls). Two diagnostic groups were established: 44 children at risk of reading difficulties (including ADHD or DLD) and 49 typically developing controls. Participants were also classified by academic cycle: 55 in the first cycle (1st–2nd grade) and 38 in the second cycle or higher (3rd–6th grade). Linguistic variables were measured through tests administered individually, with data collected during one-on-one assessment sessions. Results: Among the variables analysed, significant differences were observed only in morphological skills and OMA. No significant differences were found based on sex, whereas both academic cycle and diagnostic group showed significant effects. Conclusion: The most relevant and novel finding is that the type and frequency of errors in the OMA task could serve as an early indicator of students at risk of RD. OMA assessment could therefore be a promising method of early screening and targeted interventions. Full article

Background: Given the central role of writing and typing in contemporary communication, integrating writing assessments into clinical practice is crucial for improving the diagnosis and management of primary progressive aphasia (PPA). This scoping review summarizes evidence on writing abilities in PPA, examining task types, their strengths and limitations, the linguistic features of stimuli, and the influence of language differences. Methods: A literature search was conducted using the Google Scholar and PubMed databases. We included papers published in peer-reviewed journals and written in English that present data from at least one PPA subject and report a quantitative score relative to a writing task. Fifty-one studies were included (forty-seven behavioral; four with neuroimaging). Results: Overall, the literature is fragmented, with marked variability in task design and the control of psycholinguistic variables. Writing to dictation is the most frequently used task but fails to capture the full spectrum of writing impairments, whereas tasks tapping lexico-semantic, morpho-syntactic, and discourse-level abilities are rarely employed. At the syndromic description level, svPPA typically shows surface dysgraphia, nfvPPA presents phonological dysgraphia and agrammatic writing, and lvPPA displays mixed error profiles. Neuroimaging findings are highly heterogeneous. Conclusions: The review underscores the need for systematic, linguistically grounded approaches to writing assessments in PPA to enhance diagnostic precision and cross-linguistic comparability. Full article

Background/Objectives: Phonemic restoration refers to improved speech understanding when periodic silent interruptions are replaced by a plausible masking sound, reflecting an interaction between perceptual continuity and top-down linguistic inference. This study tested whether the magnitude and rate dependence of phonemic restoration vary systematically with stimulus complexity, operationalized using speech materials that differ in response constraints and linguistic variability. Methods: Young adults with normal audiometric thresholds completed an interrupted-speech identification task using five corpora spanning closed-set and open-set speech corpora. Stimuli were periodically interrupted at 2 Hz and 3 Hz with a 50% duty cycle. For each corpus and rate, interruption intervals were either left silent or filled with speech-shaped noise. Results: Closed-set materials yielded higher intelligibility than open-set materials across conditions. Replacing silent gaps with speech-shaped noise improved intelligibility for all corpora. Importantly, the joint influence of interruption rate and gap-filler depended on the stimulus type: rate-by-filler interactions were most evident for the open-set corpora as compared to the closed-set corpora. Keyword identification varied systematically with word position for the open-set materials, indicating nonuniform vulnerability across sentence structures. Conclusions: These results indicate that phonemic restoration is robust but material-dependent. Stimulus complexity shapes how temporal sampling and masking plausibility combine to support perceptual repair, and open-set, high-variability materials are particularly sensitive to these interactions. Full article

Background/Objectives: In recent years, research on psychosis has increasingly focused on prevention, aiming to implement early interventions that mitigate or reduce its impact. Within this framework, the analysis of linguistic markers in individuals with at-risk mental states (ARMS) has proven valuable for identifying those at risk and predicting psychosis onset. Artificial intelligence tools, particularly natural language processing (NLP), have emerged as effective resources for detecting these language-based indicators. This study aims to synthesize the existing scientific evidence on linguistic markers analyzed through NLP techniques in individuals with ARMS. Methods: A systematic review following the PRISMA 2020 protocol was conducted. Three databases (PubMed, PsycInfo, and Scopus) were searched for published articles from their inception to October 2025. Rayyan software was used to manage references and article downloads. Out of ninety initial search results, fifteen studies involving 1313 participants from diverse groups were included in the review. Results: The findings indicated that alterations in semantic coherence, syntactic complexity, referential cohesion, and speech/content poverty differentiated ARMS individuals from healthy controls. Several of these markers, analyzed with NLP methods, predicted the onset of psychosis with accuracy levels ranging from 79% to 100%, although these findings should be interpreted with caution due to the significant methodological heterogeneity and variability in sample sizes across the included studies. Conclusions: NLP techniques offer a powerful approach for detecting language alterations that distinguish ARMS individuals and provide meaningful predictions of psychosis onset, highlighting their potential as a complement to traditional clinical assessments for early identification and prevention. Full article

Understanding how users perceive and attend to the anthropomorphic appearance and empathic responses of artificial intelligence medical conversational agents (AIMCAs) can help reveal the key judgment cues underlying trust formation and use decisions, while also informing interface and dialog design. To this end, this study employs a 3 (appearance anthropomorphism: high, medium, low) × 2 (empathic response: present, absent) within-subject eye-tracking experiment, combined with subjective scales and brief post-task open-ended feedback. During a static prototype viewing task based on hypothetical consultation scenarios, we concurrently recorded trust, behavioral intention, and visual measures for key areas of interest (AOIs; appearance area, conversational content area, and overall interface area). Eye-tracking measures were normalized by AOI coverage proportion to improve cross-AOI comparability. The results show that both anthropomorphic appearance and empathic response significantly increased users’ trust in AIMCAs and their behavioral intention. An interaction between these two types of social cues was also observed, suggesting that when visual embodiment and linguistic style are aligned at the social level, users are more likely to form favorable overall judgments. At the level of visual processing, however, no interaction effect was found, and the eye-tracking measures showed only partial main effects, indicating that subjective synergy does not necessarily correspond to synergistic changes in attentional allocation. Overall, anthropomorphic appearance and empathic response exerted consistent facilitating effects on outcome variables, but displayed different patterns of attentional allocation and information prioritization at the visual level. Accordingly, AIMCA design should emphasize consistency between appearance cues and conversational strategies, optimize users’ initial judgments and interface comprehension, and use intention through verifiable information organization and clear boundary cues. Full article

Speaking proficiency remains one of the most challenging skills for learners of English as a Foreign Language (EFL), particularly in contexts where sustained spoken interaction is limited. This systematic review synthesises 36 empirical studies (2015–2025) identified through a PRISMA-guided Scopus search to examine how artificial intelligence (AI)-mediated instruction supports EFL speaking development. The included studies were analysed according to AI modality, pedagogical integration, instructional input characteristics, and linguistic and affective outcomes. Findings indicate that AI tools—such as chatbots, automatic speech recognition systems, and large language models—consistently support affective outcomes, including reduced speaking anxiety and increased willingness to communicate. Improvements in fluency, pronunciation, and accuracy were frequently reported, particularly when AI tools were embedded within task-based and pedagogically structured instructional designs. However, evidence for sustained development of higher-order communicative competence was more variable. The review proposes a mediated input framework conceptualising AI as a design-sensitive instructional resource rather than an autonomous teaching agent. Full article

With the rapid development of intensive marine aquaculture, water quality has become a key factor affecting both economic benefits and ecological safety in marine aquaculture. In the process of actual water quality evaluation, due to the great uncertainty and ambiguity of evaluation indicators, experts find it difficult to evaluate in real number form and are more inclined to use linguistic variables to evaluate indicators, which poses challenges for the construction of water quality evaluation models. An intuitionistic fuzzy set (IFS) is an effective tool for dealing with uncertainty and fuzziness in complex problems. Based on a detailed analysis of existing distance measures for IFS, this study proposes a new distance measure that not only considers membership and non-membership information, but also constructs an allocation function for membership and non-membership, introducing hesitation information into distance metrics. We proposed the definitions and proved the properties. The comparative experiments show that the new distance measure can overcome the shortcomings of existing distance measures. Furthermore, based on the newly proposed distance measure, the IFS TOPSIS method is improved in multi-attribute decision-making applications. Finally, a practical application of marine aquaculture water quality evaluation is used. The results illustrate that when $\alpha$ = 1 the closeness declines from 0.741 to 0.432, when =2 the closeness declines from 0.662 to 0.46, and when =6 the closeness declines from 0.566 to 0.82. The convenience and effectiveness of the new method is demonstrated. Full article

Despite concern over air pollution, fine-scale spatial and demographic disparities in exposure remain largely unquantified in Australian cities due to sparse monitoring and coarse models. In Greater Melbourne, this gap limits neighbourhood-level assessment of PM_2.5 exposure and associated environmental inequalities. To address this gap, we integrated 6-month averaged PM_2.5 observations (October 2023 to March 2024) from 5 regulatory monitoring stations and 13 low-cost sensors (LCSs) to develop a land use regression (LUR) model estimating concentrations at a 100 m resolution. These estimates were used to calculate population-weighted PM_2.5 exposure (PWE) at the mesh block level across Melbourne. To examine factors associated with spatial heterogeneity in PWE, we applied a hybrid modeling framework combining Spatially Explicit Random Forest (Spatial-RF) and Geographically Weighted Regression (GWR), incorporating physical, built-environment, and socio-demographic variables from the Synthesized Multi-Dimensional Environmental Exposure Database (SEED). The Spatial-RF model initially exhibited an R² of 0.56. After multicollinearity diagnostics using the Variance Inflation Factor (VIF), three key explanatory variables were selected for GWR modeling: the Normalized Difference Vegetation Index (NDVI), the Index of Education and Occupation (IEO), and the proportion of culturally and linguistically diverse populations (CALDP). The developed GWR model achieved higher model performance (R² = 0.65) than Spatial-RF and global Ordinary Least Squares (OLS) regression (R² = 0.38), revealing strong spatial non-stationarity. Results show that PWE generally ranged from 5 to 7 µg/m³, exceeding the 2021 WHO air quality guideline, with hotspots in the urban core and along major transport corridors. Elevated exposure occurred in both socioeconomically disadvantaged areas and residents in urban centers with higher socio-economic status, reflecting complex, spatially contingent exposure inequalities. These findings support fine-scale, equity-oriented air quality management. Full article

This paper reports on a pilot study evaluating five large language models (ChatGPT-4, Gemini 2.0 Flash, Claude 3.5 Sonnet, Perplexity AI, and DeepSeek) in gradient acceptability judgment tasks in German. The models rated 150 contextually embedded sentences on a 5-point Likert scale across five categories: gray-zone (variable) items, canonical grammatical items, ungrammatical items, diatopically marked items, and diastratically/diaphasically marked items. All models clearly distinguish between clearly grammatical and clearly ungrammatical stimuli in unambiguous morphosyntactic contexts. Mixed-effects analyses further show that differences between models vary across stimulus categories rather than reflecting a uniform global shift in acceptability ratings. These findings indicate that current LLMs robustly capture core morphosyntactic contrasts, but that model behavior is less uniform in domains involving variation and contextual sensitivity. The study contributes to the empirical assessment of LLMs as acceptability raters and informs debates on their methodological role in linguistics. Full article

This study reconstructs the geography of meaning of the German perception verb schmecken on the basis of 30 major dialect dictionaries, treating them as a distributed semantic corpus and coding attestations as binary variables reflecting the presence or absence of semantic options. Combining a construal-based framework with spatial modeling, the analysis shows that the polysemy of schmecken is structured by three mutually reinforcing forces: embodied sensory organization, construal-based perspectivization, and regionally patterned areal dynamics. The gustatory–olfactory axis forms the semantic core of the verb, from which tactile, visual, affective, and epistemic extensions emerge. These extensions align with systematic pathways constrained by agentive, experiential, emissive, and evaluative construals, demonstrating that semantic extension is channeled through specific construal modes—notably emissive and agentive—rather than determined by sensory modality alone. A detailed areal analysis reveals a pronounced north–south divide. While Low German dialects conform to the cross-linguistically more common tendency to avoid colexifying taste and smekk—itself the outcome of historical change rather than uninterrupted differentiation—Upper German varieties preserve a typologically rare gustatory–olfactory cluster and exhibit the richest range of cross-modal and abstract extensions. The resulting semantic graph formalizes how regional varieties activate different subsets of a lexeme’s semantic potential and demonstrates that semantic networks themselves display spatial organization. The study thus provides an empirically grounded reconstruction of a German geography of meaning and illustrates how dialect data illuminate the interplay between embodied cognition, construal-based lexical architecture, and areal dynamics. Full article

Multi-class and multi-label classification of medical dialogues remains a challenging task due to high linguistic variability and transcription noise. This study proposes an ensemble approach based on three fine-tuned Polish T5 (Text-to-Text Transfer Transformer) models trained on partially overlapping clinical dialogue datasets. The models are evaluated exclusively on low-quality, highly noisy, automatically transcribed conversations to assess real-world robustness. The results demonstrate that the ensemble of models improves classification stability and outperforms the best single model, increasing the F1-score by 21.8% for internal medicine dialogues and by 44.9% for paediatric interviews. The proposed method shows potential for practical deployment in clinical decision support and automated medical documentation systems. Full article