Search Results (678)

With social media as the main channel for college students to express emotions, sentiment public opinion analysis in big data environments poses three core challenges to campus sentiment monitoring and psychological counseling: severe data noise interference, insufficient multi-attribute feature extraction, and the trade-off between recognition accuracy and inference efficiency. This paper proposes a university student public opinion prediction model integrating multi-attribute decision-making and BERT–Mamba. First, an anti-interference matching filter cleans raw data by filtering out advertisements and irrelevant comments to improve data quality. Second, a multi-attribute decision object model extracts quantifiable attributes covering media sources, themes, and temporal dimensions. Third, BERT generates textual sentiment representations, and a three-stage deep feature extraction architecture with Mamba balances accuracy and efficiency. Finally, multi-attribute features and sentiment representations are fused for dynamic public opinion prediction. Validated using the ChnSentiCorp Chinese sentiment analysis benchmark dataset and university student Weibo public opinion corpus, the model achieves 97.44% average sentiment recognition accuracy. It provides technical support for universities to understand student sentiment trends and address negative public opinions, with practical value for enhancing campus public opinion monitoring and assisting mental health counseling. Full article

Antimicrobial resistance (AMR) in aquaculture environments has become an increasing concern due to the dissemination of antibiotic-resistant bacteria in aquatic ecosystems. This study investigated the occurrence of presumptive antimicrobial-resistant bacteria in commercial fish farms in Hajdú-Bihar county, Hungary. Between 2023 and 2024, water samples were collected from fishponds from Hajdúszoboszló, Biharugra, and Debrecen, followed by membrane filtration with 0.45 µm pore size mixed cellulose ester membrane filters, bacterial isolation using selective media including vancomycin-resistant enterococci (VRE) agar, Eosin Methylene Blue (EMB) agar, and Endo’s medium. The phenotypic characterization revealed the presence of diverse bacterial populations that could be associated with potential antimicrobial resistance. Bacterial identification using Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (MALDI-TOF MS) showed several bacterial genera, predominantly Acinetobacter, Pseudomonas, Enterococcus, Bacillus, and Escherichia. A total of 16 presumptive antimicrobial-resistant bacterial isolates recovered from selective media were identified using MALDI-TOF MS. Several detected species, including Acinetobacter baumannii, Pseudomonas aeruginosa, Enterococcus faecium, and Enterococcus faecalis, are recognized as opportunistic bacteria associated with antimicrobial resistance dissemination. The findings demonstrate that aquaculture waters in Hajdú-Bihar county may serve as reservoirs of antimicrobial-resistant bacteria. Full article

Groundwater in northeastern China is characterized by low temperature and the coexistence of manganese (Mn), ammoniacal nitrogen, and organic pollutants, resulting in prolonged start-up of biological Mn removal filters. To address this challenge, combined bioaugmentation with heterotrophic nitrifying bacteria (HNB) and manganese-oxidizing bacteria (MnOB) was employed. The optimal inoculation ratio was determined, and the effects of manganese sand and activated carbon on Mn and ammoniacal nitrogen removal were evaluated. Under bioaugmentation conditions, manganese sand promoted the rapid establishment of ammonia oxidation, whereas activated carbon exhibited superior Mn²⁺ oxidation performance. A dual-media filter consisting of an upper manganese sand layer and a lower activated carbon layer was subsequently developed. Compared with the conventional manganese sand filter, the bioaugmented composite filter shortened the start-up period by 21 days and reached stable operation on day 24. High-throughput 16S rRNA gene sequencing revealed that activated carbon enriched norank_f__Beggiatoaceae and enhanced synergistic interactions among functional microbial groups. This study provides a promising strategy for the efficient simultaneous removal of Mn and ammoniacal nitrogen from groundwater in cold regions. Full article

The influence of pleat angle on the pressure drop of H14 HEPA filters was investigated through a mathematical model that represents the filter as a system of converging–diverging channels coupled with porous filtration media. The analysis was conducted for pleat angles ranging from 1° to 20° under a constant laminar airflow rate of 0.167 m³/s and 0.45 m/s velocity. The model combines Darcy–Forchheimer flow through the filtration media with laminar channel flow theory, enabling the total pressure drop to be expressed as a function of pleat geometry and subsequently optimised through analytical differentiation. The results show that the pressure drop contribution of the filtration media increases with the pleat angle, from 10.57 Pa at 1° to 213.49 Pa at 20°, whereas channel losses decrease sharply from 1121.71 Pa to 2.75 Pa over the same interval. The competing behaviour of these two mechanisms generates a minimum total pressure drop of 94.56 Pa at a pleat angle of approximately 6°, compared with 120 Pa for the current industrial configuration operating at 3.73°. This represents a pressure drop reduction of approximately 21.2%, implying a corresponding decrease in fan energy consumption without compromising filtration performance. The analysis further demonstrates that very small pleat angles (1–2°) are highly unfavourable, producing total pressure drops between 301 and 1132 Pa due to severe channel constriction, while for angles above 13–14°, the channel contribution becomes negligible, and the overall pressure drop is governed almost entirely by the filtration media. These findings provide quantitative design criteria for optimising HEPA, EPA, and ULPA filter geometries, highlighting pleat angle as a critical parameter for improving aerodynamic performance, flow uniformity, and energy efficiency in high-purity environments. The proposed model was further assessed using a commercially available H14 HEPA filter with 188 pleats, an effective filtration area of 10.618 m², and a nominal airflow rate of 600 m³/h, demonstrating its applicability to real industrial filter configurations. Full article

The broad adoption of Large Language Models (LLMs) has increased the need for human-curated datasets that serve as evaluation benchmarks. This need is particularly pronounced for non-English languages and for tasks that are inherently subjective and require multiple human perspectives. One such example is the development of benchmarks designed to assess the cultural awareness of LLMs. Statistics and data science courses offer a potential setting for developing such benchmarks while teaching students to apply LLM evaluation techniques using statistical inference. This paper presents a pilot project in which students in a statistics course within a data science engineering program created culturally diverse multiple-choice questions, generated answers using LLMs, and applied statistical methods to assess model accuracy. Student feedback indicated the project was engaging and useful for learning, while also highlighting a notable reliance on LLMs, particularly for interpreting statistical results. The resulting dataset comprises 1355 multiple-choice questions across 18 categories, including language, social media, and politics. After filtering valid items, the dataset was used to evaluate both closed- and open-source LLMs. Results show that the Gemini (closed-source) and Qwen (open-source) model families achieved the best performance, with improvements linked to model size, reasoning capabilities, and access to search tools. The best closed-source model achieved an accuracy of 97.66%, whereas the best open-source model achieved an accuracy of 79.07%. Qualitative analyses of errors in the filtering procedure and model reasoning process point to possible explanations into the challenges LLMs face when handling culturally specific content. Furthermore, results support a cultural injection hypothesis, whereby cultural knowledge is embedded during pretraining and accessed through instruction tuning. Through this work, we aim to demonstrate how statistics and data science courses can provide productive contexts for developing open-source benchmarks for non-English languages while also enriching students’ learning experiences. The dataset is publicly available. Full article

Photoacoustic signal analysis in weakly absorbing media remains challenging because of low signal-to-noise ratios. This work proposes a deep learning framework for classifying particle size and concentration in an indirect absorption configuration. We conducted a comparative study using raw temporal signals, Savitzky–Golay filtering, and time–frequency scalograms via Continuous Wavelet Transform (CWT), and evaluated both 1D and 2D convolutional neural network architectures. Experimental validation was performed using poly(methyl methacrylate) (PMMA) microspheres (6 μm and 15 μm) and hydroxyapatite nanoparticles (<200 nm) at volume fractions as low as $6\times {10}^{-4}\%$. While raw signals led to unstable training (accuracy ≈ 47%), CWT-based representations significantly improved performance, achieving near-perfect size discrimination and over 96% accuracy in discrete volume-fraction classification. Grad-CAM analysis confirmed that the model identifies physically meaningful regions of the acoustic waveform, ensuring interpretability. The proposed framework was validated under controlled experimental conditions using discrete particle types and predefined volume-fraction classes, providing a foundation for future extensions toward continuous particle characterization. Ultimately, these findings demonstrate that combining time–frequency representations with deep learning provides a robust, physically consistent approach for particle characterization in turbid media, with significant potential for biomedical diagnostics and material analysis. Full article

Indoor air pollution (IAP) is an important public health issue in urban households, particularly in hot climates where residents spend substantial time indoors. While the health risks of indoor air pollution are increasingly recognised, gender differences in exposure-related behaviours, risk perception, and health communication preferences remain underexplored in Gulf cities. This study examined gender differences in awareness, household practices, and preferred information channels related to IAP among English-speaking adults in Dubai. A cross-sectional survey was conducted among 1014 adults using a structured questionnaire. Descriptive, bivariate, and multivariable regression analyses were used to examine gender-based patterns. Overall, participants showed moderate awareness of IAP; gender differences emerged across each of the four behavioural domains examined, with women more likely to use chemical-based cleaners and to open windows for ventilation. After controlling for age and education, women were significantly more likely than men to use traditional chemical-based cleaners (β = 0.69; p < 0.001) and to open windows for ventilation (β = 0.78; p < 0.001), while men more frequently relied on mechanical methods such as air filters and ceiling fans. Women were more likely to express interest in learning about IAP (adjusted OR = 2.55; 95% CI: 1.71–3.79) and to prefer community-based, informal channels such as awareness campaigns and social media. These gender differences persisted after adjustment for age and education, suggesting that indoor air quality behaviours and communication preferences may vary systematically by gender in this sample. Among this English-speaking convenience sample, gender differences were observed in household practices and preferred communication channels. These findings provide preliminary evidence that may help inform future public health communication strategies in Dubai but should not be interpreted as representative of the wider Dubai population. Full article

Participatory citizenship plays a critical role in strengthening climate change resilience, particularly in the context of natural disasters such as floods and other hydrological extremes. Citizen-generated data shared through social media platforms offer valuable real-time insights that can complement traditional environmental monitoring systems. This study proposes a machine learning-based framework to analyze multilingual news data and global X (formerly known as Twitter) data that can complement street level sensor data for improved detection and understanding of extreme hydrological events: floods and landslides. The approach identifies and filters tweets related to hazards such as floods and contextualizes them with information extracted from news reports to enhance event characterization. In addition, sentiment and emotion analysis are applied to assess public reactions and perceived event intensity. By integrating physical event signals with societal responses, the method provides a broader perspective on disaster impacts and the effectiveness of emergency responses. The results highlight the potential of combining social media analytics and machine learning to support hydrological monitoring, enhance situational awareness, and contribute to more responsive disaster management strategies in the face of increasing climate-related risks. Full article

Luminescent sensing in aqueous media using organic small-molecule emitters is often constrained by water-induced fluorescence quenching and indicator leakage. In this study, a polymer-confined thermally activated delayed fluorescence (TADF) material, poly-BrTPPz, was synthesized by covalently copolymerizing a donor–acceptor monomer into a polyacrylamide network. Density functional theory calculations indicate spatial frontier orbital separation. The polymer matrix restricts intramolecular motion, while the polar amide microenvironment provides a solid-state solvation effect, decreasing the reverse intersystem crossing barrier to activate delayed luminescence with a lifetime of 303 μs and a photoluminescence quantum yield of 69.1% in the solid state. In aqueous environments, the material exhibits a selective quenching response toward iron(III) ions (Fe³⁺) through a mechanism involving the inner filter effect and pyrazine-coordinated static quenching. To mitigate potential secondary environmental contamination, a transmembrane diffusion model was evaluated by encapsulating the polymer within a semi-permeable membrane, which limits indicator leakage while permitting analyte permeation. This work outlines a design approach for environment-responsive luminescent devices in closed aquatic systems. Full article

Roman Urdu hate and offensive language detection is difficult because informal social-media posts combine non-standard spelling, code-mixing, local abusive expressions, and severe label imbalance. Existing transformer-based detectors improve contextual modeling, but they often leave lexical threat evidence implicit and rarely address whether model confidence is reliable when large language model teacher signals are used in low-resource training. We propose Dynamic Gated Distillation with Symbolic Evidence (DGDS), a reliability-gated neuro-symbolic learning framework for Roman Urdu hate and offensive language detection. DGDS combines a cross-lingual semantic encoder, a context-modulated symbolic threat profiler, and a reliability-gated dual-stream learner. The key idea is to treat the large language model teacher output as a noisy structured soft-label record rather than as a replacement for gold labels. DGDS therefore uses the teacher probability assigned to the gold class, the current student confidence, and a teacher reliability gate to form a dynamic importance weight for supervised training. Under a leakage-aware evaluation protocol on the RUHSOLD benchmark, the central contribution of DGDS is confidence reliability: it holds macro-F1 at parity with the strongest dual-stream baseline on both the binary and five-class settings while substantially reducing the expected calibration error from 0.0512 to 0.0134 on Task 1 (a 74% relative reduction) and from 0.0813 to 0.0268 on Task 2 (a 67% relative reduction). These calibration gains are statistically significant ($p<0.001$, paired stratified bootstrap) and robust to teacher replacement (GPT-4o-mini yields ECE 0.0158 vs. the 0.0512 baseline on Task 1). These results support a conservative conclusion: large language model soft labels are most useful for Roman Urdu harmful-content detection when they are filtered as uncertain training evidence and deployed through a local calibrated student model. Full article

This study examines how the chromatic characteristics of the Min River urban waterfront are represented in publicly circulated social media images to support place-based biophilic landscape design. Dominant-color records were extracted through pixel-level filtering and per-image K-means clustering in CIELAB space, with image-level lighting-condition interpretation and record-level landscape-element labeling assisted by a multimodal large language model and subsequently reviewed and corrected by the author. Hue distributions and five record-proportion-weighted metrics of saturation, value, vividness, chromatic dissonance, and complexity were examined through an analytical framework integrating temporal scenarios, landscape elements, and lighting conditions. The results reveal a recurrent blue–orange orientation produced by the complementary positioning of multiple landscape elements rather than by any single category. Nighttime imagery showed the highest saturation but the lowest value, whereas dawn and dusk combined relatively high saturation and value and produced the highest vividness and chromatic dissonance. Transitional illumination brought built surfaces closer to natural elements in saturation–value space, while Ward hierarchical clustering identified Color-Affinity Groups that crossed temporal, lighting-condition, and landscape-element boundaries. These findings support a relational interpretation of biophilic color as a condition-dependent configuration rather than a fixed set of element-bound hues. Weighted representative palettes provide scenario-sensitive design references. However, the findings characterize publicly circulated visual representations of the waterfront rather than calibrated physical-color measurements or direct evidence of restorative effects. Full article

Fake social media profiles increasingly resemble ordinary accounts, combining believable images, fluent biographies, copied posts, and selective engagement, which makes single signal detection unreliable. This study develops and evaluates a forensic natural language processing (NLP) framework that treats profile text as digital evidence rather than raw model input. Profiles were collected from Facebook, Instagram, X, Truth Social, and LinkedIn through network expansion and vocabulary guided scraping, then anonymized, merged, cleaned, and organized into structured text and profile records. Ground truth came from human eye annotation (referred to throughout as human eye review, human review, or manual review) supported by three LLM reviewers (referred to throughout as LLM assisted review), OpenAI GPT-4o, Anthropic Claude 3.5 Sonnet, and Google Gemini 1.5 Pro, with strict agreement logic producing confidence-tiered labels across platform-specific label spaces covering scam, fraud, harmful, and legitimate behavior. Each model received a defined set of label categories, was instructed to return a single label with a one sentence rationale, and worked independently with no cross-model communication. Multidimensional features, including TF-IDF representations, sentence embeddings, scam lexicon counts, sentiment and emotion scores, and behavioral indicators, were extracted through a pipeline built to keep links, hashtags, contact markers, and repeated phrasing as forensic signals rather than noise. Traditional classifiers, transformer models, and ensembles were trained on verified labels, and the best model per platform was applied to the full dataset of over 309,108 records under confidence filtered classification, with predictions below a 0.70 probability threshold flagged for human review. The framework links classification outputs to textual indicators through explainable AI, achieving weighted F1 scores between 0.86 and 0.96 across five platforms. These results show that profile text alone provides strong, production-relevant evidence for detecting fake and harmful accounts, even though the framework treats this as one evidence stream to be weighed alongside image and behavioral signals rather than as a final determination on its own. The accompanying evidence trail means an investigator or a trust and safety team can see why a profile was flagged rather than being handed a bare score, which is what turns a classifier into a tool that can be checked, challenged, and used directly in an investigation. Full article

Spam detection on SMS messaging has not received as much attention from researchers recently as the spam detection studies on emails or social media platforms. However, spam SMS messaging can be more intrusive, annoying, and harmful. Thus, detecting and filtering spam SMS messages is becoming a priority that saves human productivity. This work aims to introduce a dynamic, accurate, and efficient machine learning-based spam detection technique for SMS messaging. It aims at protecting users and businesses from spam SMS attacks. It aims to detect and identify suspicious messages that contain promotional, misleading, irrelevant, or harmful content. It primarily aims to test and evaluate the impact of feature extraction methods on the performance of machine-learning-based spam detection. Several text feature extraction techniques have been used and tested, including classical, statistical, contextual, and advanced embedding techniques. An extensive set of experiments has been presented on benchmark datasets in this field. From the comparative study, we can infer that all investigated feature extraction techniques have achieved high accuracy (90%+) on the in-domain dataset. However, their performance decreased when they were tested on the out-of-domain dataset (70%+). The advanced embedding techniques achieved the best performance across both datasets compared to the other tested feature extraction models. Full article

Fake news detection has become a major research topic at the intersection of artificial intelligence, data mining, and information security. In this paper, we evaluate the performance of English-trained algorithms on English translations of Romanian-sourced news articles, using a translation-mediated cross-domain evaluation design. The study is based on source code generated with the assistance of artificial intelligence systems for a set of machine learning and transformer-based models. The code was subsequently implemented in Google Colab. 2026, trained on international benchmark datasets, and tested on Romanian news content. This design allowed the rapid prototyping of multiple detection pipelines and the systematic observation of their behavior in a media environment different from that represented in the training corpora. The models were evaluated comparatively using standard classification metrics, including accuracy, precision, recall, and F1-score, complemented by additional indicators relevant to model robustness and practical usability. The experimental results revealed significant differences in performance across algorithms when applied to English translations of Romanian-language news content after training on international datasets. However, this study does not provide a direct comparison between model performance on the international benchmark datasets and the Romanian test corpus; therefore, the gap between the international training corpus and the Romanian-sourced test corpus is interpreted as an exploratory limitation and as a direction for future research. Based on these findings, we propose an empirical classification of the tested models according to their predictive effectiveness, their contextual robustness across linguistic environments, and their operational relevance as filtering tools for institutional monitoring. The results show that AI-assisted coding workflows can provide a viable starting point for reproducible misinformation research, but they also underline the limitations of directly transferring models trained on non-Romanian data to local media ecosystems. The study offers both a replicable evaluation framework and practical insights for institutions involved in strategic communication, public security, and the monitoring of information threats. Full article

The treatment of domestic wastewater remains a significant environmental challenge, particularly in rural areas where the construction of sewer systems is often not economically viable. Vertical-flow filters offer an alternative, but traditional filter media are prone to clogging, which justifies the search for new waste-based fillers. The aim of this study was to evaluate the potential of synthetic waste wool as a filter medium and biofilm carrier in vertical filters for domestic wastewater treatment, with particular emphasis on the effects of active bed height and hydraulic loading rate. Experimental models were used to assess the influence of these parameters on treatment performance. High removal efficiencies of organic pollutants and suspended solids were achieved. The filter with an active bed height of 0.9 m that operated at a hydraulic loading rate of 80 dm³·m⁻²·d⁻¹ achieved removal efficiencies of 93.06–97.65% for BOD₅, 86.53–94.20% for COD, 62.69–87.53% for TSS, and 65.73–94.31% for NH₄⁺-N. The most favorable treatment results were obtained for a greater bed depth and lower hydraulic load values. An increase in hydraulic load resulted in a deterioration of the quality of the treated wastewater. The results obtained indicate that waste synthetic wool can serve as a promising filtration material in decentralized wastewater treatment systems, aligning with the principles of the circular economy and the sustainable development of environmental technologies. Full article