Search Results (104)

Recent advancements in large language models (LLMs) have significantly enhanced the effectiveness of Retrieval-Augmented Generation (RAG) systems. This study focuses on the development and evaluation of a domain-specific AI chatbot designed to support international student admissions by leveraging LLM-based RAG pipelines. We implement and compare multiple pipeline configurations, combining retrieval methods (e.g., Dense, MMR, Hybrid), chunking strategies (e.g., Semantic, Recursive), and both open-source and commercial LLMs. Dual evaluation datasets of LLM-generated and human-tagged QA sets are used to measure answer relevancy, faithfulness, context precision, and recall, alongside heuristic NLP metrics. Furthermore, latency analysis across different RAG stages is conducted to assess deployment feasibility in real-world educational environments. Results show that well-optimized open-source RAG pipelines can offer comparable performance to GPT-4o while maintaining scalability and cost-efficiency. These findings suggest that the proposed chatbot system can provide a practical and technically sound solution for international student services in resource-constrained academic institutions. Full article

Tourist satisfaction models typically assume that service performance dimensions carry the same weight for all travelers. Drawing on Bourdieu, we reconceptualize age, gender, and region of origin as demographic capital, durable resources that mediate how visitors decode service cues. Using a SERVPERF-based survey of 407 international travelers departing Quito (Ecuador), we test measurement invariance across six sociodemographic strata with multi-group confirmatory factor analysis. The four-factor SERVPERF core (Access, Lodging, Extra-hotel Services, Attractions) holds, yet partial metric invariance emerges: specific loadings flex with demographic capital. Gen-Z travelers penalize transport reliability and safety; female visitors reward cleanliness and empathy; and Latin American guests are the most critical of basic organization. These patterns expose a boundary condition for universalistic satisfaction models and elevate demographic capital from a descriptive tag to a structuring construct. Managerially, we translate the findings into segment-sensitive levers, visible security for youth and regional markets, gender-responsive facility upgrades, and dual eco-luxury versus digital-detox bundles for long-haul segments. By demonstrating when and how SERVPERF fractures across sociodemographic lines, this study intervenes in three theoretical conversations: (1) capital-based readings of consumption, (2) the search for boundary conditions in service-quality measurement, and (3) the shift from segmentation to capital-sensitive interpretation in emerging markets. The results position Ecuador as a critical case and provide a template for destinations facing similar performance–perception mismatches in the Global South. Full article

The accelerating pace of digital life has intensified psychological strain, increasing the demand for accessible and systematized emotional support tools. Relaxing video games—defined as low-pressure, non-competitive games designed to promote calm and emotional relief—offer immersive environments that facilitate affective engagement and sustained user involvement. This study proposes a computational framework that integrates sentiment analysis and topic modeling to investigate the affective mechanisms and behavioral dynamics associated with relaxing gameplay. We analyzed nearly 60,000 user reviews from the Steam platform in both English and Chinese, employing a hybrid methodology that combines sentiment classification, dual-stage Latent Dirichlet Allocation (LDA), and multi-label mechanism tagging. Emotional relief emerged as the dominant sentiment (62.8%), whereas anxiety was less prevalent (10.4%). Topic modeling revealed key affective dimensions such as pastoral immersion and cozy routine. Regression analysis demonstrated that mechanisms like emotional relief (β = 0.0461, p = 0.001) and escapism (β = 0.1820, p < 0.001) were significant predictors of longer playtime, while Anxiety Expression lost statistical significance (p = 0.124) when contextual controls were added. The findings highlight the potential of relaxing video games as scalable emotional regulation tools and demonstrate how sentiment- and topic-driven modeling can support system-level understanding of affective user behavior. This research contributes to affective computing, digital mental health, and the design of emotionally aware interactive systems. Full article

Proteins regulate various cellular processes and are of great biological interest. The protein search engine is a crucial tool in proteomics research, used to analyze high-throughput tandem mass spectrometry data and to identify protein sequence information. A core step in protein search engines is constructing sequence tag indexes and performing the rapid retrieval of protein databases. However, as the scale of protein sequence data continues to grow, traditional protein search engines face the dual challenges of the high storage cost of sequence tag indexes and low retrieval efficiency. To address these issues, we propose a sequence tag index scheme named STIP, which is based on an inverted index and compression techniques. Based on STIP, we design a peptide retrieval algorithm named STIP-Search. This algorithm utilizes the sequence tag index constructed by STIP for peptide sequence retrieval. STIP uses the greedy algorithm to partition the tag index into blocks; in this way, STIP can generate tag indexes for very large protein databases, such as NCBI-nr. Compared to the current four mainstream tag index generation algorithms used in Open-pFind, MODplus, TIIP and PIPI2, STIP has the lowest storage and time consumption. It utilizes delta encoding, index reduction, and dynamic bit width encoding to compress the tag index, reducing the storage cost by 76.2%. Compared to TIIP, which is currently the algorithm with the lowest time complexity, the time cost of the peptide sequence retrieval of STIP-Search is reduced by 8.94% to 23.31%. Full article

Nanoquenchers with a single quenching cofactor exhibit limited fluorescence quenching efficiency. In this work, a metal–organic hybrid with dual quenching cofactors (Cu²⁺ and pyrroloquinoline quinone or PQQ) was prepared by metal-coordinated assembly and used as a nanoquencher for a protease assay with enhanced quenching efficiency. The peptide substrate with an oligohistidine (His₆) tag was labeled with a fluorophore. Caspase-3 was determined as a protease example. The substrate was attached onto the surface of the Cu-PQQ nanoquencher by a metal coordination interaction between the unsaturated Cu²⁺ on the nanoparticle surface and the His₆ tag in the peptide. The cleavage of the peptide substrate by enzymatic hydrolysis led to the release of a fluorophore-conjugated segment from the nanoquencher surface, thus turning on the fluorescence. The nanoprobe was used to determine caspase-3 with a linear range of 0.01–5 ng/mL and a detection limit of 7 pg/mL. Furthermore, the method was used to evaluate inhibition efficiency and monitor drug-induced cell apoptosis. In contrast to other means of peptide immobilization, such as physical adsorption and covalent coupling, the strategy based on the metal coordination interaction is simple and powerful, thereby achieving assays of caspase-3 activity in lysates with a satisfactory result. The work should be valuable for the design of nanoquenchers with multiple quenching cofactors and the development of novel biosensors. Full article

This paper seeks to address the limitations of conventional remote sensing data dissemination algorithms, particularly their inability to model fine-grained multi-modal heterogeneous feature correlations and adapt to dynamic network topologies under resource constraints. This paper proposes multi-modal-MAPPO, a novel multi-modal deep reinforcement learning (MDRL) framework designed for a proactive data push in large-scale integrated satellite–terrestrial networks (ISTNs). By integrating satellite cache states, user cache states, and multi-modal data attributes (including imagery, metadata, and temporal request patterns) into a unified Markov decision process (MDP), our approach pioneers the application of the multi-actor-attention-critic with parameter sharing (MAPPO) algorithm to ISTNs push tasks. Central to this framework is a dual-branch actor network architecture that dynamically fuses heterogeneous modalities: a lightweight MobileNet-v3-small backbone extracts semantic features from remote sensing imagery, while parallel branches—a multi-layer perceptron (MLP) for static attributes (e.g., payload specifications, geolocation tags) and a long short-term memory (LSTM) network for temporal user cache patterns—jointly model contextual and historical dependencies. A dynamically weighted attention mechanism further adapts modality-specific contributions to enhance cross-modal correlation modeling in complex, time-varying scenarios. To mitigate the curse of dimensionality in high-dimensional action spaces, we introduce a multi-dimensional discretization strategy that decomposes decisions into hierarchical sub-policies, balancing computational efficiency and decision granularity. Comprehensive experiments against state-of-the-art baselines (MAPPO, MAAC) demonstrate that multi-modal-MAPPO reduces the average content delivery latency by 53.55% and 29.55%, respectively, while improving push hit rates by 0.1718 and 0.4248. These results establish the framework as a scalable and adaptive solution for real-time intelligent data services in next-generation ISTNs, addressing critical challenges in resource-constrained, dynamic satellite–terrestrial environments. Full article

Background/Objectives: Biodegradable polymers have emerged as promising platforms for drug delivery. Produced by microbiomes, polyhydroxyalkanoates (PHAs) offer excellent biocompatibility, biodegradability, and environmental sustainability. In this study, we report the surface functionalization of PHA-based nanoparticles (NPs) using the SpyTag–SpyCatcher system to enhance cellular uptake. Methods: Initial conjugation with mEGFP-SpyTag enabled visualization, followed by decoration with HER2-specific Affibody-SpyCatcher and/or TAT-SpyCatcher peptides. The prepared NPs retained a diameter of <200 nm and a negatively charged surface. Results: Affibody-functionalized NPs significantly enhanced internalization and cytotoxicity in HER2-overexpressing SK-BR-3 cells, whereas TAT-functionalized NPs promoted uptake across various cell types, independently of HER2 expression. Dual-functionalized NPs exhibited synergistic or attenuated effects based on the HER2 expression levels, highlighting the critical role of ligand composition in targeted delivery. Conclusions: The results of this study demonstrate that the SpyTag–SpyCatcher-mediated surface engineering of PHA NPs offers a modular and robust strategy for active targeting in nanomedicine. Full article

Accurate parameter estimation is essential for effective monitoring and treatment of high-salinity industrial wastewater. Traditional methods such as spectroscopy, ion chromatography, and electrochemical analysis offer high sensitivity but are often complex, costly, and unsuitable for real-time monitoring. This research integrates Deep Neural Networks (DNNs) with the Levenberg–Marquardt (LM) algorithm to develop an advanced RFID-based sensing system for real-time monitoring of sodium chloride solutions in wastewater. The DNN extracts essential features from raw data, while the LM algorithm optimizes parameter estimation for enhanced precision and stability. Experimental results show that the dielectric constant sample variance at various flow rates under wireless frequency is 0.08509, while the sample total variance is 0.06807, both below 0.1. Additionally, the sample standard deviation and total standard deviation are both below 0.3, at 0.26090 and 0.29169, respectively. These findings confirm that the proposed system is robust against flow rate variations, ensuring accurate, real-time monitoring and supporting sustainable industrial practices. Full article

This paper presents a dual circularly polarized encoding terahertz tag utilizing linked-semi-ellipses elements. The proposed design achieves broadband 2-bit polarization encoding in terahertz frequency bands through independently controlling dual circularly polarized phases with linked-semi-ellipses elements. The tag simultaneously possesses two key functionalities: circular polarization conversion and circular polarization encoding, which enhance its resilience to environmental influences. Simulations demonstrated the polarization conversion rates exceeding 80% across 0.42–0.53 THz with reflected RCS from tag achieved −51.7 dB·m² and a 3 dB beamwidth of ±22°. The tag achieves broadband 2-bit circular polarization encoding and attains a 23.6% relative bandwidth. The circular polarization encoding method enhances stability in metal-rich environments and offers substantial potential for diverse practical applications, including indoor positioning, logistics tracking, and object identification. Full article

Radio frequency identification (RFID) technology has gained significant attention in various industry sectors due to its potential for efficient inventory management, asset tracking, and object localization. Different RFID technologies are available; resorting to harmonic signals is currently less used but, recently, has gained interest in research activity. In this study, we present the design, prototype realization, and performance evaluation of a dual-band dual-polarized harmonic tag. The tag incorporates a dual-band matching circuit utilizing a zero-bias Schottky diode HSMS-2850 connected to a perturbed annular ring patch antenna. The antenna, in fact, is able to radiate in cross-polarization at the higher frequency. Through a comprehensive design methodology, including simulation optimization and prototype fabrication, we demonstrate the successful implementation of the tag. Measurements to validate the impedance matching properties, radiation patterns, and backscattering capability of the tag are also shown. Full article

Tetrahydrofolate (THF), the biologically active form of folate, serves as a crucial carrier of one-carbon units essential for synthesizing cellular components such as amino acids and purine nucleotides in vivo. It also acts as an important precursor for the production of pharmaceuticals, including folinate and L-5-methyltetrahydrofolate (L-5-MTHF). In this study, we developed an efficient enzyme cascade system for the production tetrahydrofolate from folate, incorporating NADPH recycling, and explored its application in the synthesis of L-5-MTHF, a derivative of tetrahydrofolate. To achieve this, we first screened dihydrofolate reductases (DHFRs) from various organisms, identifying SmDHFR from Serratia marcescens as the enzyme with the highest catalytic activity. We then conducted a comparative analysis of formate dehydrogenases (FDHs) from different sources, successfully establishing an NADPH recycling system. To further enhance biocatalytic efficiency, we optimized key reaction parameters, including temperature, pH, enzyme ratio, and substrate concentration. To address the challenge of pH mismatch in dual-enzyme reactions, we employed an enzymatic microenvironment regulation strategy. This involved covalently conjugating SmDHFR with a superfolder green fluorescent protein mutant carrying 30 surface negative charges (−30sfGFP), using the SpyCatcher/SpyTag system. This modification resulted in a 2.16-fold increase in tetrahydrofolate production, achieving a final yield of 4223.4 µM. Finally, we extended the application of this tetrahydrofolate synthesis system to establish an enzyme cascade for L-5-MTHF production with NADH recycling. By incorporating methylenetetrahydrofolate reductase (MTHFR), we successfully produced 389.8 μM of L-5-MTHF from folate and formaldehyde. This work provides a novel and efficient pathway for the biosynthesis of L-5-MTHF and highlights the potential of enzyme cascade systems in the production of tetrahydrofolate-derived compounds. Full article

Radio Frequency Identification (RFID) is a key component in automatic systems that address challenges in environment monitoring. However, tag collision continues to be an essential challenge in such applications due to high-density RFID deployments. This paper addresses the issue of RFID tag collision in large-scale intensive tags, particularly in industrial membrane contamination monitoring systems, and improves the system performance by minimizing collision rates through an innovative collision-avoiding algorithm. This research improved the Predictive Framed Slotted ALOHA–Collision Tracking Tree (PRFSCT) algorithm by cooperating probabilistic and deterministic methods through dynamic frame length adjustment and multi-branch tree processes. After simulation and validation in MATLAB R2023a, we performed a hardware test with the RFM3200 and UHFReader18 passive tags. The method’s efficiency is evaluated through collision slot reduction, delay minimization, and enhanced throughput. PRFSCT significantly reduces collision slots when the number of tags to identify is the same for PRFSCT, Framed Slotted ALOHA (FSA), and Collision Tracking Tree (CTT); the PRFSCT method needs the fewest time slots. When identifying more than 200 tags, PRFSCT has 225 collision slots for 500 tags compared to FSA and CTT, which have approximately 715 and 883 for 500 tags, respectively. It demonstrates exceptional stability and adaptability under increased density needs while improving tag reading at distances. Full article

Audio tagging, as a fundamental task in acoustic signal processing, has demonstrated significant advances and broad applications in recent years. Spiking Neural Networks (SNNs), inspired by biological neural systems, exploit event-driven computing paradigms and temporal information processing, enabling superior energy efficiency. Despite the increasing adoption of SNNs, the potential of event-driven encoding mechanisms for audio tagging remains largely unexplored. This work presents a pioneering investigation into event-driven encoding strategies for SNN-based audio tagging. We propose the SATRN (Spiking Audio Tagging Robust Network), a novel architecture that integrates temporal–spatial attention mechanisms with membrane potential residual connections. The network employs a dual-stream structure combining global feature fusion and local feature extraction through inverted bottleneck blocks, specifically designed for efficient audio processing. Furthermore, we introduce an event-based encoding approach that enhances the resilience of Spiking Neural Networks to disturbances while maintaining performance. Our experimental results on the Urbansound8k and FSD50K datasets demonstrate that the SATRN achieves comparable performance to traditional Convolutional Neural Networks (CNNs) while requiring significantly less computation time and showing superior robustness against noise perturbations, making it particularly suitable for edge computing scenarios and real-time audio processing applications. Full article

The red palm weevil (RPW; Coleoptera: Curculionidae) is a destructive pest affecting palms worldwide, capable of causing significant economic losses and ecological damage in managed palm ecosystems. Current management heavily relies on synthetic insecticides, but their overuse fosters resistance. Bacillus thuringiensis (Bt) offers a promising alternative, producing toxins selective against various insect orders, including Coleoptera. However, no specific Bt toxin has yet been identified for RPW. This study investigates the toxicity against RPW larvae of the Bt Cry1Ia protoxin, known for its dual activity against Lepidoptera and Coleoptera. A laboratory RPW colony was reared for two generations, ensuring a reliable insect source for bioassays. Cry1Ia was expressed as a 6xHis-tagged fusion protein in Escherichia coli and purified using nickel affinity. Incubation with RPW larval gut proteases for 24 h produced a stable core of ~65 kDa. Diet-incorporation bioassays revealed high Cry1Ia toxicity in neonate larvae. In contrast, the lepidopteran-active Cry1Ac protoxin, used as a robust negative control, was completely degraded after 24 h of in vitro proteolysis and showed no toxicity in bioassays. Cry1Ia-fed larvae exhibited significant midgut cell damage, characteristic of Bt intoxication. These findings highlight Cry1Ia’s strong potential for integration into RPW management programs. Full article

The development and regulation of hair are widely influenced by biological rhythm signals. Melatonin plays a crucial role as a messenger in transmitting biological rhythm signals, and its impact on hair development has been well documented. During the process of hair follicle reconstruction, hair follicle stem cells (HFSCs) are the most important cell type, but the regulatory effect of melatonin on the state of HFSCs is still not fully understood. Therefore, it is necessary to conduct a more comprehensive characterization of the effects of melatonin on the state of hair follicle stem cells. The research results indicate that HFSCs express retinoic acid receptor-related orphan receptor alpha (Rorα), and melatonin inhibits the expression level of RORA. Experimental results from CUT&Tag, CUT&RUN, and dual luciferase reporter assays demonstrate that Foxc1 is a downstream target gene of RORA, with RORA regulating Foxc1 expression by binding to the promoter region of Foxc1. The CCK-8 assay results show that low doses of melatonin upregulate the survival rate of hair follicle stem cells, while high doses have the opposite effect. The knockdown of Foxc1 reverses the inhibitory effect of high-dose melatonin on the survival rate of hair follicle stem cells. Based on these findings, we believe that melatonin-mediated circadian signals exert a bidirectional regulatory effect on the state of HFSCs. Full article