Search Results (509)

Adversarial attacks in Natural Language Processing (NLP) present a critical challenge, particularly in sentiment analysis, where subtle input modifications can significantly alter model predictions. In search of more robust defenses against adversarial attacks on sentimental analysis, this research work introduces two novel defense mechanisms: the Lexicon-Based Random Substitute Model (LRSM) and the Word-Variant Voting Model (WVVM). LRSM employs randomized substitutions from a dataset-specific lexicon to generate diverse input variations, disrupting adversarial strategies by introducing unpredictability. Unlike traditional defenses requiring synonym dictionaries or precomputed semantic relationships, LRSM directly substitutes words with random lexicon alternatives, reducing overhead while maintaining robustness. Notably, LRSM not only neutralizes adversarial perturbations but occasionally surpasses the original accuracy by correcting inherent model misclassifications. Building on LRSM, WVVM integrates LRSM, Frequency-Guided Word Substitution (FGWS), and Synonym Random Substitution and Voting (RS&V) in an ensemble framework that adaptively combines their outputs. Logistic Regression (LR) emerged as the optimal ensemble configuration, leveraging its regularization parameters to balance the contributions of individual defenses. WVVM consistently outperformed standalone defenses, demonstrating superior restored accuracy and F1 scores across adversarial scenarios. The proposed defenses were evaluated on two well-known sentiment analysis benchmarks: the IMDB Sentiment Dataset and the Yelp Polarity Dataset. The IMDB dataset, comprising 50,000 labeled movie reviews, and the Yelp Polarity dataset, containing labeled business reviews, provided diverse linguistic challenges for assessing adversarial robustness. Both datasets were tested using 4000 adversarial examples generated by established attacks, including Probability Weighted Word Saliency, TextFooler, and BERT-based Adversarial Examples. WVVM and LRSM demonstrated superior performance in restoring accuracy and F1 scores across both datasets, with WVVM excelling through its ensemble learning framework. LRSM improved restored accuracy from 75.66% to 83.7% when compared to the second-best individual model, RS&V, while the Support Vector Classifier WVVM variation further improved restored accuracy to 93.17%. Logistic Regression WVVM achieved an F1 score of 86.26% compared to 76.80% for RS&V. These findings establish LRSM and WVVM as robust frameworks for defending against adversarial text attacks in sentiment analysis. Full article

Soil salinization severely limits plant growth and productivity. Mulberry (Morus alba L.), an economically and ecologically important tree, is widely cultivated, yet its salt-tolerance mechanisms at the seedling stage remain insufficiently understood. This study investigated the physiological and biochemical responses of two-year-old mulberry (‘Tailai Sang’) seedlings subjected to six NaCl treatments (0, 50, 100, 150, 200, and 300 mmol L⁻¹) for 28 days. Results showed that growth parameters and photosynthetic gas exchange exhibited dose-dependent declines. The reduction in net photosynthetic rate (Pn) was attributed to both stomatal limitations (decreased stomatal conductance) and non-stomatal limitations, as evidenced by a significant decrease in the maximum quantum efficiency of photosystem II (Fv/Fm) under high salinity. To cope with osmotic stress, seedlings accumulated compatible solutes, including soluble sugars, proteins, and proline. Critically, mulberry seedlings demonstrated effective ion homeostasis by sequestering Na⁺ in the roots to maintain a high K⁺/Na⁺ ratio in leaves, a mechanism that was compromised above 150 mmol L⁻¹. Concurrently, indicators of oxidative stress—malondialdehyde (MDA) and H₂O₂—rose significantly with salinity, inducing the activities of antioxidant enzymes (SOD, CAT, APX, and GR), which peaked at 150 mmol L⁻¹ before declining under extreme stress. A biomass-based LC₅₀ of 179 mmol L⁻¹ NaCl was determined. These findings elucidate that mulberry salt tolerance is a coordinated process involving three key mechanisms: osmotic adjustment, selective ion distribution, and a robust antioxidant defense system. This study establishes an indicative tolerance threshold under controlled conditions and provides a physiological basis for further field-based evaluations of ‘Tailai Sang’ mulberry for cultivation on saline soils. Full article

The integration of digital twins (DTs) with intelligent traffic systems (ITSs) holds strong potential for improving real-time management in smart cities. However, securing digital twins remains a significant challenge due to the dynamic and adversarial nature of cyber–physical environments. In this work, we propose TwinFedPot, an innovative digital twin-based security architecture that combines honeypot-driven data collection with Zero-Shot Learning (ZSL) for robust and adaptive cyber threat detection without requiring prior sampling. The framework leverages Inverse Federated Distillation (IFD) to train the DT server, where edge-deployed honeypots generate semantic predictions of anomalous behavior and upload soft logits instead of raw data. Unlike conventional federated approaches, TwinFedPot reverses the typical knowledge flow by distilling collective intelligence from the honeypots into a central teacher model hosted on the DT. This inversion allows the system to learn generalized attack patterns using only limited data, while preserving privacy and enhancing robustness. Experimental results demonstrate significant improvements in accuracy and F1-score, establishing TwinFedPot as a scalable and effective defense solution for smart traffic infrastructures. Full article

This study investigated the immunonutritional potential of high-molecular-weight (Mw~85 kDa), non-degraded chitosan (NCS) and gamma-radiation-degraded, low-molecular-weight chitosan (RCS) incorporated into aquafeeds for Nile tilapia (Oreochromis niloticus). RCS was produced by γ-irradiation (10 kGy) in the presence of 0.25% (w/v) H₂O₂, yielding low-viscosity, colloidally stable nanoparticles with Mw ranging from 10 to 13 kDa. Five diets were formulated: a control, NCS at 0.50%, and RCS at 0.025%, 0.050%, and 0.075%. No adverse effects on growth were observed, confirming safety. Immune gene expression (e.g., ifng1, nfκb, tnf), antioxidant markers (e.g., reduced MDA, increased GSH and GR), and nonspecific humoral responses (lysozyme, IgM, and bactericidal activity) were significantly enhanced in the NCS-0.50, RCS-0.050, and RCS-0.075 groups. Notably, these benefits were achieved with RCS at 10-fold lower concentrations than NCS. Following challenge with Edwardsiella tarda, fish fed RCS-0.050 and RCS-0.075 diets exhibited the highest survival rates and relative percent survival, highlighting robust activation of innate and adaptive immunity alongside redox defense. These results support the use of low-Mw RCS as a biologically potent, cost-effective alternative to traditional high-Mw chitosan in functional aquafeeds. RCS-0.050 and RCS-0.075 show strong potential as immunonutritional agents to enhance fish health and disease resistance in aquaculture. Full article

As cyberattacks continue to rise in frequency and sophistication, extracting actionable Cyber Threat Intelligence (CTI) from diverse online sources has become critical for proactive threat detection and defense. However, accurately identifying complex entities from lengthy and heterogeneous threat reports remains challenging due to long-range dependencies and domain-specific terminology. To address this, we propose XLNet-CRF, a hybrid framework that combines permutation-based language modeling with structured prediction using Conditional Random Fields (CRF) to enhance Named Entity Recognition (NER) in cybersecurity contexts. XLNet-CRF directly addresses key challenges in CTI-NER by modeling bidirectional dependencies and capturing non-contiguous semantic patterns more effectively than traditional approaches. Comprehensive evaluations on two benchmark cybersecurity corpora validate the efficacy of our approach. On the CTI-Reports dataset, XLNet-CRF achieves a precision of 97.41% and an F1-score of 97.43%; on MalwareTextDB, it attains a precision of 85.33% and an F1-score of 88.65%—significantly surpassing strong BERT-based baselines in both accuracy and robustness. Full article

Novel networking paradigms such as the Internet of Things (IoT) have expanded their usage and deployment to various application domains. Consequently, unseen critical security vulnerabilities such as zero-day attacks have emerged in such deployments. The design of intrusion detection systems for IoT networks is often challenged by a lack of labeled data, which complicates the development of robust defenses against adversarial attacks. As deep learning-based network intrusion detection systems, network intrusion detection systems (NIDS) have been used to counteract emerging security vulnerabilities. However, the deep learning models used in such NIDS are vulnerable to adversarial examples. Adversarial examples are specifically engineered samples tailored to a specific deep learning model; they are developed by minimal perturbation of network packet features, and are intended to cause misclassification. Such examples can bypass NIDS or enable the rejection of regular network traffic. Research in the adversarial example detection domain has yielded several prominent methods; however, most of those methods involve computationally expensive retraining steps and require access to labeled data, which are often lacking in IoT network deployments. In this paper, we propose an unsupervised method for detecting adversarial examples that performs early detection based on the intrinsic characteristics of the deep learning model. Our proposed method requires neither computationally expensive retraining nor extra hardware overhead for implementation. For the work in this paper, we first perform adversarial example generation on a deep learning model using autoencoders. After successful adversarial example generation, we perform adversarial example detection using the intrinsic characteristics of the layers in the deep learning model. A robustness analysis of our approach reveals that an attacker can easily bypass the detection mechanism by using low-magnitude log-normal Gaussian noise. Furthermore, we also test the robustness of our detection method against further compromise by the attacker. We tested our approach on the Kitsune datasets, which are state-of-the-art datasets obtained from deployed IoT network scenarios. Our experimental results show an average adversarial example generation time of 0.337 s and an average detection rate of almost 100%. The robustness analysis of our detection method reveals a reduction of almost 100% in adversarial example detection after compromise by the attacker. Full article

Aging in sexually reproducing organisms is shaped by the declining force of natural selection after reproduction begins. In Drosophila melanogaster, experimental evolution shows that altering the age of reproduction shifts the timing of aging. Using the Drosophila experimental evolution population (DEEP) resource, which includes long- and short- lived populations evolved under distinct reproductive schedules, we investigated how immune defense against Beauveria bassiana changes with age and evolved lifespan. We tested survival post-infection at multiple ages and examined genomic differentiation for immune-related genes. Both population types showed age-related declines in immune defense. Long-lived populations consistently exhibited age-specific defense when both long- and short-lived populations were tested. Genomic comparisons revealed thousands of differentiated loci, yet no enrichment for canonical immune genes or overlap with gene sets from studies of direct selection for immunity. These results suggest that enhanced immune defense can evolve alongside extended lifespan, likely via general physiological robustness rather than traditional immune pathways. A more detailed analysis may reveal that selection for lifespan favors tolerance-based mechanisms that reduce infection damage without triggering immune activation, in contrast to direct selection for resistance. Our findings demonstrate the utility of experimentally evolved populations for dissecting the genetic architecture of aging and immune defense to inform strategies to mitigate age-related costs associated with immune activation. Full article

In the face of escalating cyber threats to critical infrastructures, achieving robust cyber resilience has become paramount. This paper proposes an endogenous security-oriented framework for cyber resilience assessment, specifically tailored for critical infrastructures. Drawing on the principles of endogenous security, our framework integrates dynamic heterogeneous redundancy (DHR) and adaptive defense mechanisms to address both known and unknown threats. We model resilience across four key dimensions—Prevention, Destruction Resistance, Adaptive Recovery, and Evolutionary Learning—using a novel mathematical formulation that captures nonlinear interactions and temporal dynamics. The framework incorporates environmental threat entropy to dynamically adjust resilience scores, ensuring relevance in evolving attack landscapes. Through empirical validation on simulated critical infrastructure scenarios, we demonstrate the framework’s ability to quantify resilience trajectories and trigger timely defensive adaptations. Empiricalvalidation on a real-world critical infrastructure system yielded an overall resilience score of 82.75, revealing a critical imbalance between strong preventive capabilities (90/100) and weak Adaptive Recovery (66/100). Our approach offers a significant advancement over static risk assessment models by providing actionable metrics for strategic resilience investments. This work contributes to the field by bridging endogenous security theory with practical resilience engineering, paving the way for more robust protection of critical systems against sophisticated cyber threats. Full article

Automatic Speaker Verification (ASV) systems are increasingly employed to secure access to services and facilities. However, recent advances in speech deepfake generation pose serious threats to their reliability. Modern speech synthesis models can convincingly imitate a target speaker’s voice and generate realistic synthetic audio, potentially enabling unauthorized access through ASV systems. To counter these threats, forensic detectors have been developed to distinguish between real and fake speech. Although these models achieve strong performance, their deep learning nature makes them susceptible to adversarial attacks, i.e., carefully crafted, imperceptible perturbations in the audio signal that make the model unable to classify correctly. In this paper, we explore adversarial attacks targeting speech deepfake detectors. Specifically, we analyze the effectiveness of Basic Iterative Method (BIM) attacks applied in both time and frequency domains under white- and black-box conditions. Additionally, we propose an ensemble-based attack strategy designed to simultaneously target multiple detection models. This approach generates adversarial examples with balanced effectiveness across the ensemble, enhancing transferability to unseen models. Our experimental results show that, although crafting universally transferable attacks remains challenging, it is possible to fool state-of-the-art detectors using minimal, imperceptible perturbations, highlighting the need for more robust defenses in speech deepfake detection. Full article

Background/Objectives: Endocrine disruptors (EDs) are xenobiotic chemicals that disrupt hormone signaling and homeostasis within the human body. Accumulative evidence proposes that EDs could affect systemic hormone balance and local microbial communities, including the female genital tract (FGT) microbiome. The FGT microbiome, and especially the vaginal microbiota, contributes significantly to reproductive health maintenance, defense against infection, and favorable pregnancy outcomes. Disruption of the delicate microbial environment is associated with conditions like bacterial vaginosis, infertility, and preterm birth. Methods: The present narrative review summarizes the existing literature on EDs’ potential for changing the FGT microbiome. We discuss EDs like bisphenol A (BPA), phthalates, and parabens and their potential for disrupting the FGT microbiome through ED-induced hormone perturbations, immune modulation, and epithelial barrier breach, which could lead to microbial dysbiosis. Results: Preliminary evidence suggests that ED exposure–microbial composition changes relationships; however, robust human evidence for EDs’ changes on the FGT microbiome remains scarce. Conclusions: Our review addresses major research gaps and suggests future directions for investigation, such as the necessity for longitudinal and mechanistic studies that combine microbiome, exposome, and endocrine parameters. The relationship between EDs and the FGT microbiome could be critical for enhancing women’s reproductive health and for steering regulatory policies on exposure to environmental chemicals. Full article

Viral infections persistently challenge global health through immune evasion and zoonotic transmission. Dendritic cells (DCs) play a central role in antiviral immunity by detecting viral nucleic acids via conserved pattern recognition receptors, triggering interferon-driven innate responses and cross-presentation-mediated activation of cytotoxic CD8⁺ T cells. This study synthesizes DC-centric defense mechanisms against viral subversion, encompassing divergent nucleic acid sensing pathways for zoonotic DNA and RNA viruses, viral counterstrategies targeting DC maturation and interferon signaling, and functional specialization of DC subsets in immune coordination. Despite advances in DC-based vaccine platforms, clinical translation is hindered by cellular heterogeneity, immunosuppressive microenvironments, and limitations in antigen delivery. Future research should aim to enhance the efficiency of DC-mediated immunity, thereby establishing a robust scientific foundation for the development of next-generation vaccines and antiviral therapies. A more in-depth exploration of DC functions and regulatory mechanisms may unlock novel strategies for antiviral intervention, ultimately paving the way for improved prevention and treatment of viral infections. Full article

Losses from floods and the wide range of impacts have been at the forefront of hazard-triggered disasters in China. Affected by large-scale human activities and the environmental evolution, China’s defense flood situation is undergoing significant changes. This paper constructs a comprehensive flood disaster risk assessment model through systematic analysis of four key factors—hazard (H), exposure (E), susceptibility/sensitivity (S), and disaster prevention capabilities (C)—and establishes an evaluation index system. Using the Analytic Hierarchy Process (AHP), we determined indicator weights and quantified flood risk via the following formula R = H × E × V × C. After we applied this model to 16 towns in coastal Zhejiang Province, the results reveal three distinct risk tiers: low (R < 0.04), medium (0.04 ≤ R ≤ 0.1), and high (R > 0.1). High-risk areas (e.g., Longxi and Shitang towns) are primarily constrained by natural hazards and socioeconomic vulnerability, while low-risk towns benefit from a robust disaster mitigation capacity. Risk typology analysis further classifies towns into natural, social–structural, capacity-driven, or mixed profiles, providing granular insights for targeted flood management. The spatial risk distribution offers a scientific basis for optimizing flood control planning and resource allocation in the district. Full article

Mobility data plays a crucial role in modern location-based services (LBSs), yet it poses significant privacy risks, as it can reveal highly sensitive information such as home locations and behavioral patterns. This paper focuses on the anonymization of mobility data by obfuscating mobility heat maps and combining this with a local differential privacy method, which generates synthetic mobility traces. Using the San Francisco Cabspotting dataset, we compare the effectiveness of the combined approach against reidentification attacks. Our results show that mobility traces treated with both a heat map obfuscation and local differential privacy are less likely to be reidentified than those anonymized solely with Heat Map Confusion. This two-tiered anonymization process balances the trade-off between privacy and data utility, providing a robust defense against reidentification while preserving data accuracy for practical applications. The findings suggest that the integration of synthetic trace generation with heat map-based obfuscation can significantly enhance the protection of mobility data, offering a stronger solution for privacy-preserving data sharing. Full article

Background/Objectives: Artificial intelligence (AI) is beginning to be applied in wound ballistics, showing preliminary potential to improve the accuracy and objectivity of forensic analyses. This review explores the current state of AI applications in forensic firearm wound analysis, emphasizing its potential to address challenges such as subjective interpretations and data heterogeneity. Methods: A systematic review adhering to PRISMA guidelines was conducted using databases such as Scopus and Web of Science. Keywords focused on AI and GSW classification identified 502 studies, narrowed down to 4 relevant articles after rigorous screening based on inclusion and exclusion criteria. Results: These studies examined the role of deep learning (DL) models in classifying GSWs by type, shooting distance, and entry or exit characteristics. The key findings demonstrated that DL models like TinyResNet, ResNet152, and ConvNext Tiny achieved accuracy ranging from 87.99% to 98%. Models were effective in tasks such as classifying GSWs and estimating shooting distances. However, most studies were exploratory in nature, with small sample sizes and, in some cases, reliance on animal models, which limits generalizability to real-world forensic scenarios. Conclusions: Comparisons with other forensic AI applications revealed that large, diverse datasets significantly enhance model performance. Transparent and interpretable AI systems utilizing techniques are essential for judicial acceptance and ethical compliance. Despite the encouraging results, the field remains in an early stage of development. Limitations highlight the need for standardized protocols, cross-institutional collaboration, and the integration of multimodal data for robust forensic AI systems. Future research should focus on overcoming current data and validation constraints, ensuring the ethical use of human forensic data, and developing AI tools that are scientifically sound and legally defensible. Full article

As a pivotal model organism in Lepidoptera research, the silkworm (Bombyx mori) holds significant importance in life science due to its economic value and biotechnological applications. Advancements in proteomics and bioinformatics have enabled substantial progress in characterizing the B. mori proteome. Systematic screening and identification of protein–protein interactions (PPIs) have progressively elucidated the molecular mechanisms governing key biological processes, including viral infection, immune regulation, and growth development. This review comprehensively summarizes traditional PPI detection techniques, such as yeast two-hybrid (Y2H) and immunoprecipitation (IP), alongside emerging methodologies such as mass spectrometry-based interactomics and artificial intelligence (AI)-driven PPI prediction. We critically analyze the strengths, limitations, and technological integration strategies for each approach, highlighting current field challenges. Furthermore, we elaborate on the molecular regulatory networks of Bombyx mori nucleopolyhedrovirus (BmNPV) from multiple perspectives: apoptosis and cell cycle regulation; viral protein invasion and trafficking; non-coding RNA-mediated modulation; metabolic reprogramming; and host immune evasion. These insights reveal the dynamic interplay between viral replication and host defense mechanisms. Collectively, this synthesis aims to provide a robust theoretical foundation and technical guidance for silkworm genetic improvement, infectious disease management, and the advancement of related biotechnological applications. Full article