Search Results (130)

The collaborative construction of large-scale, diverse datasets is crucial for developing high-performance machine learning models. However, this collaboration faces significant challenges, including ensuring data security, protecting participant privacy, maintaining high dataset quality, and aligning economic incentives among multiple stakeholders. Effective risk management strategies are essential to systematically identify, assess, and mitigate potential risks associated with data collaboration. This study proposes a federated blockchain-based framework designed to manage multiparty dataset collaborations securely and transparently, explicitly incorporating comprehensive risk management practices. The proposed framework involves six core entities—key distribution center (KDC), researcher (RA), data owner (DO), consortium blockchain, dataset evaluation platform, and the orchestrating model itself—to ensure secure, privacy-preserving and high-quality dataset collaboration. In addition, the framework uses blockchain technology to guarantee the traceability and immutability of data transactions, integrating token-based incentives to encourage data contributors to provide high-quality datasets. To systematically mitigate dataset quality risks, we introduced an innovative categorical dataset quality assessment method leveraging label reordering to robustly evaluate datasets. We validated this quality assessment approach using both publicly available (UCI) and privately constructed datasets. Furthermore, our research implemented the proposed blockchain-based management system within a consortium blockchain infrastructure, benchmarking its performance against existing methods to demonstrate enhanced security, reliability, risk mitigation effectiveness, and incentive alignment in dataset collaboration. Full article

Nowadays, consumers are becoming increasingly concerned about the husbandry conditions under which animals are raised, particularly broilers, since broilers are one of the species whose welfare is most impaired in intensive farming systems. One of the primary means of communicating husbandry practices to consumers is through product labelling. Thus, a rising number of animal welfare labelling schemes for broilers are being developed and used across Europe by initiatives of both public and private stakeholders, including NGOs that advocate for animal welfare. This review aims to identify, analyze, and compare these labelling schemes with a focus on the main animal welfare provisions included in them, which contribute to enhanced animal welfare. The schemes were identified through web searches, so that we could visit their official websites, access their standards and regulations and study them in detail. We included in our research only those schemes whose criteria were publicly available. In total, 16 schemes were selected and analyzed. Although these schemes vary in their criteria, they all enhance the welfare standards of broiler production, primarily through incorporating environmental enrichment and/or access to the outdoors. Most schemes define and specify in detail the required animal welfare provisions, setting a clear application frame for the raising period of the birds. However, the welfare of animals during transport and slaughter is often overlooked. Full article

With the dramatic growth in dataset size, active learning has become one of the effective methods to deal with large-scale unlabeled data. However, most of the existing active learning methods are inefficient due poor target models and lack the ability to utilize the feature similarity between labeled and unlabeled data. Furthermore, data leakage is a serious threat to data privacy. In this paper, considering the features of the data itself, an augmented graph convolutional network is proposed which acts as a sampler for data selection in active learning, avoiding the involvement of the initial poor target model. Then, by applying the proposed GCN as a substitute for the initial poor target model, this paper proposes an active learning model based on augmented GCNs, which is able to select more representative data, enabling the active learning model to achieve better classification performance with limited labeled data. Finally, this paper proposes a homomorphic encryption-based federated active learning model to improve the data utilization and enhance the security of private data. Experiments were conducted on three datasets, Cora, CiteSeer and PubMed, and achieved accuracy rates of 94.47%, 92.86% and 91.51%, respectively, while providing provable security guarantees. Furthermore, the highest malicious user detection accuracy was 88.07%, and the global model test accuracy reached 88.42%, 84.22% and 81.46%, under a model poisoning attack. Full article

With the rapid development of the digital economy and the evolving shopping preferences of consumers, e-commerce platforms have been enhancing their competitiveness by launching private label brands and optimizing their sales channel strategies. This study focuses on an online sales system comprising a strong brand and an e-commerce platform. Four game modes were constructed: agency selling only (NN), agency selling combined with reselling (NS), agency selling combined with private labels (IN), and reselling combined with agency selling under the introduction of private labels (IS). Under the coupon promotion strategy, this study focused on the introduction strategy for private labels (PLs) and the selection strategy for platform sales modes. Our research produced the following findings: (1) Regardless of whether the platform introduces its own brand, adopting a reselling mode can significantly enhance the profits of both the brand owner and platform. (2) Irrespective of whether the reselling mode is implemented, the platform’s profits are always increased when introducing its own brand. (3) When the coupon redemption rate is higher, the brand owner achieves greater profitability in the absence of PL introduction. Conversely, when the coupon redemption rate is low, an increase in the commission rate leads to reduced profit margins for the brand owner due to competition from a private label. (4) When the coupon redemption and commission rate are both high, the coupon face value without a PL is larger. Otherwise, when these rates are both low, the coupon face value is higher under the introduction of a PL. This study offers a theoretical foundation and decision-making support for e-commerce platforms to optimize sales mode selection, introduce private-label brands, and develop coupon strategies. Full article

In the Netherlands, around 52,000 dairy cows die on the primary farm each year due to natural death, euthanasia, or on-farm emergency slaughter (OFES). The decision as to what is the best option is made by the farmer, often after consulting a veterinarian, a livestock trader, or a slaughterhouse operator. To determine which factors play a role in this decision-making process, semi-structured interviews were conducted with dairy farmers, private veterinary practitioners, livestock traders, and slaughterhouse operators in the Netherlands. Dairy cattle culling decisions are influenced and limited by strict enforcement of livestock transport regulations and limited options for on-farm killing methods. Requirements regarding mortality rates imposed by the dairy industry and private quality labels for raw milk also influence culling decisions in the Netherlands. Most participants stated that restrictive conditions regarding OFES and mobile slaughterhouses (MSHs) appear to have (unintended) negative effects on cow welfare and meat salvage in general. Different interests, such as cow welfare, food safety, economic concerns of various stakeholders, the reputational interests of the dairy and beef industries, and sustainability objectives such as meat salvage can be conflictive. The results of this study show that the decision-making process regarding culling or (prolonged) veterinary treatment of dairy cattle is complex because various factors, interests, and uncertainties must be weighed. This weighing can vary between individual dairy farms and individual dairy farmers. Full article

Background: Health consciousness refers to an individual’s level of knowledge and concern regarding the impact of food on personal health; sensory appeal to the influence of attributes such as taste, aroma, appearance, and texture on food preference; and perception of front-of-package (FOP) labels refers to how the presentation of nutritional information on the package affects product choice. Given the increasing concerns about unhealthy food consumption among university students and the role of FOP labels in guiding food choices, it is essential to understand how these factors influence purchase intentions. Objective: This study was to examine the relationship between health consciousness, sensory appeal, and perception of FOP labels with purchase intentions for unhealthy foods in university students. Methods: A cross-sectional predictive study involved 361 students from public and private universities using a non-probability purpose-sampling approach. Data were collected through a previously validated questionnaire and analyzed using multiple linear regression. Results: The results revealed a significant positive association between sensory appeal and purchase intentions for unhealthy foods (β = 0.339; p < 0.001). In contrast, health consciousness (β = −0.296; p < 0.001) and perception of FOP labels (β = −0.237; p < 0.001) were inversely related to purchase intentions. Conclusion: These findings suggest that promoting health consciousness, improving perceptions of FOP labels, and addressing sensory appeal could effectively encourage healthier eating habits and prevent diet-related diseases among university students. Full article

The human voice is an important medium of communication and expression of feelings or thoughts. Disruption in the regulatory systems of the human voice can be analyzed and used as a diagnostic tool, labeling voice as a potential “biomarker”. Conversational artificial intelligence is at the core of voice-powered technologies, enabling intelligent interactions between machines. Due to its richness and availability, voice can be leveraged for predictive analytics and enhanced healthcare insights. Utilizing this idea, we reviewed artificial intelligence (AI) models that have executed vocal analysis and their outcomes. Recordings undergo extraction of useful vocal features to be analyzed by neural networks and machine learning models. Studies reveal machine learning models to be superior to spectral analysis in dynamically combining the huge amount of data of vocal features. Clinical applications of a vocal biomarker exist in neurological diseases such as Parkinson’s, Alzheimer’s, psychological disorders, DM, CHF, CAD, aspiration, GERD, and pulmonary diseases, including COVID-19. The primary ethical challenge when incorporating voice as a diagnostic tool is that of privacy and security. To eliminate this, encryption methods exist to convert patient-identifiable vocal data into a more secure, private nature. Advancements in AI have expanded the capabilities and future potential of voice as a digital health solution. Full article

Digital transformation is increasingly relevant in food certification systems, improving processes, coordination, and data accessibility. In state-led certification systems, public entities hold a political mandate to promote digital transformation, yet little is known about digital maturity in these systems or how to assess it. This study assesses the digital maturity of a state-led food certification system in Baden-Württemberg, Germany, focusing on private sector stakeholders involved in its administration. Additionally, it examines potential measures that the governing public entity can take and evaluates the suitability of the methods used. A total of 25 out of 43 organisations were surveyed using the Digital Maturity Assessment (DMA) framework validated for the European Union (EU). Six dimensions were analysed: Digital Business Strategy, Digital Readiness, Human-Centric Digitalisation, Data Management, Automation and Artificial Intelligence, and Green Digitalisation. Data Management and Human-Centric Digitalisation were the most developed, highlighting strong data governance and workforce engagement. Automation and Artificial Intelligence were ranked lowest, reflecting minimal adoption but also indicating that not all dimensions might be of the same relevance for the variety of organisations. The variability in scores and organisation-specific relevance underscores the European DMA framework’s value, particularly due to its subsequent tailored consultation process and its integration into EU policy. Full article

In 2020, the European Union endorsed its “Farm-to-Fork” strategy, emphasizing the need for transparency in the food production chain and communication of the sustainability level and nutritional value of food products to the consumer through labeling. For animal-based products, this also includes information about the husbandry systems under which the animals are raised. At the same time, people are becoming increasingly concerned both as citizens and as consumers about animal welfare issues in production species, as animal welfare is considered an integral part of sustainability and food security. This has led to the development of various enhanced animal welfare labeling schemes, initiated by public or private entities, or even as a partnership of both. Specifically for cattle, sheep, and goats, various standards have been developed and implemented in Europe, all establishing higher welfare standards compared to conventional farming, and in some cases exceeding the minimum requirements for organic farming as set by Regulation (EU) 2018/848. Most of these standards, especially those developed by NGOs advocating for animal welfare or through public initiative, were developed for semi-intensive to extensive systems. They primarily incorporate animal-based measures, including positive welfare indicators, offering a holistic approach to animal welfare evaluation. Although there is significant heterogeneity in European animal welfare standards, nearly all of them promote access to pasture, comfort, environmental enrichment, and, in some cases, even mother–young bonding. Full article

In practical engineering, the asymmetric problem of the domain label space is inevitable owing to the prior fault information of the target domain being difficult to completely obtain. This implies that the target domain may include unseen fault classes or lack certain fault classes found in the source domain. To maintain diagnostic performance and knowledge generalization across different speeds, cross-domain intelligent fault diagnosis (IFD) models are widely researched. However, the rigid requirement for consistent domain label spaces hinders the IFD model from identifying private fault patterns in the target domain. In practical engineering, the asymmetric domain label space problem is inevitable, as the target domain’s fault prior information is difficult to completely obtain. This means that the target domain may have unseen fault classes or lack some source domain fault classes. To address these challenges, we propose an asymmetric cross-domain IFD method with label position matching and boundary sparse learning (ASY-WLB). It reduces the IFD model’s dependence on domain label space symmetry during transient speed variation. To integrate signal prior knowledge for transferable feature representation, angular resampling is used to lessen the time-varying speed fluctuations’ impact on the IFD model. We design a label-positioning information compensation mechanism and weighted contrastive domain discrepancy, accurately matching unseen class label information and constraining the diagnosis model’s decision boundary from a data conditional distribution perspective. Finally, extensive experiments on two time-varying speed datasets demonstrate our method’s superiority. Full article

Herbert D. Athearn (1923–2011) was an avid student of freshwater mollusks. He named his private shell collection “The Museum of Fluviatile Mollusks”, which was meticulously organized at his residence. This collection was curated to current museum standards with detailed labels, all lots with catalog numbers, and all unionoid valves with catalog numbers written in India ink. Specimens’ collecting dates span between 1850 and 2005, with 23,344 cataloged lots containing over 3000 lots of imperiled and extinct taxa. All data for each of the lots are handwritten in paper catalogs. Many lots contain growth series from the smallest juveniles to the largest specimens seen. He traded extensively with collectors worldwide, obtaining specimens from 84 countries. This collection was donated to the North Carolina Museum of Natural Sciences in 2007. To date, 64 percent of this collection has been databased using a relational database, totaling 589,995 specimens. The collection consists of bivalves, primarily Unionidae, Margaritiferidae, and Sphaeriidae, as well as gastropods. There are 73 families represented, with the greatest abundance found in freshwater Pleuroceridae. The Athearn collection donation included his correspondence, his library, field notes, and USGS topographic maps with marked field localities. Full article

This study investigated consumer perceptions of raw cashew nuts from two different private labels (private label A, PLA, and private label B, PLB), employing a combination of explicit (sensory analysis) and implicit (consumer neuroscience) methods. The objective was to analyse both conscious and unconscious responses to understand consumer preferences. Participants (n = 80) evaluated the samples, with electroencephalography (EEG) and electrodermal activity (EDA) as implicit methods, and hedonic scales, JAR scales, and the EsSense25 questionnaire used for explicit evaluations. The results revealed a clear preference for PLB, supported by higher global hedonic scores and a significant majority (65%) choosing PLB over PLA. EEG metrics calculated for participants’ valence, frontal alpha asymmetry (FAA) for flavour indicated greater activity in the left frontal lobe for PLB, associated with positive emotions. Task engagement (TE) measurements revealed higher engagement with PLB during flavour evaluation. Penalty analysis identified that PLA was mainly penalised for a “too weak” aroma and flavour. The EsSense25 analysis showed that cashew consumption evoked predominantly positive emotions such as “pleasant”, “satisfied”, and “calm”. In conclusion, the combination of implicit and explicit methods provided a comprehensive understanding of consumer preferences, highlighting the value of both approaches and the importance of sensory attributes in driving the overall liking of raw cashews. The findings have implications for product optimisation, market segmentation, and the development of marketing strategies in the cashew industry. Full article

Spike neural networks (SNNs) perform excellently in various domains. However, SNNs based on differential privacy (DP) protocols introduce uniform noise to the gradient parameters, which may affect the trade-off between model efficiency and personal privacy. Therefore, the adaptive differential private SNN (ADPSNN) is proposed in this work. It dynamically adjusts the privacy budget based on the correlations between the output spikes and labels. In addition, the noise is added to the gradient parameters according to the privacy budget. The ADPSNN is tested on four datasets with different spiking neurons including leaky integrated-and-firing (LIF) and integrate-and-fire (IF) models. Experimental results show that the LIF neuron model provides superior utility on the MNIST (accuracy 99.56%) and Fashion-MNIST (accuracy 92.26%) datasets, while the IF neuron model performs well on the CIFAR10 (accuracy 90.67%) and CIFAR100 (accuracy 66.10%) datasets. Compared to existing methods, the accuracy of ADPSNN is improved by 0.09% to 3.1%. The ADPSNN has many potential applications, such as image classification, health care, and intelligent driving. Full article

Traditionally, detecting misinformation (real vs. fake) and authorship (human vs. AI) have been addressed as separate classification tasks, leaving a critical gap in real-world scenarios where these challenges increasingly overlap. Motivated by this need, we introduce a unified framework—the Shared–Private Synergy Model (SPSM)—that tackles both authenticity and authorship classification under one umbrella. Our approach is tested on a novel multi-label dataset and evaluated through an exhaustive suite of methods, including traditional machine learning, stylometric feature analysis, and pretrained large language model-based classifiers. Notably, the proposed SPSM architecture incorporates multitask learning, shared–private layers, and hierarchical dependencies, achieving state-of-the-art results with over 96% accuracy for authenticity (real vs. fake) and 98% for authorship (human vs. AI). Beyond its superior performance, our approach is interpretable: stylometric analyses reveal how factors like sentence complexity and entity usage can differentiate between fake news and AI-generated text. Meanwhile, LLM-based classifiers show moderate success. Comprehensive ablation studies further highlight the impact of task-specific architectural enhancements such as shared layers and balanced task losses on boosting classification performance. Our findings underscore the effectiveness of synergistic PLM architectures for tackling complex classification tasks while offering insights into linguistic and structural markers of authenticity and attribution. This study provides a strong foundation for future research, including multimodal detection, cross-lingual expansion, and the development of lightweight, deployable models to combat misinformation in the evolving digital landscape and smart society. Full article

This paper presents the experimental evaluation and analyzes the results of the first edition of the pedestrian attribute recognition (PAR) contest, the international competition which focused on smart visual sensors based on multi-task computer vision methods for the recognition of binary and multi-class pedestrian attributes from images. The participant teams designed intelligent sensors based on vision-language models, transformers and convolutional neural networks that address the multi-label recognition problem leveraging task interdependencies to enhance model efficiency and effectiveness. Participants were provided with the MIVIA PAR Dataset, containing 105,244 annotated pedestrian images for training and validation, and their methods were evaluated on a private test set of over 20,000 images. In the paper, we analyze the smart visual sensors proposed by the participating teams, examining the results in terms of accuracy, standard deviation and confusion matrices and highlighting the correlations between design choices and performance. The results of this experimental evaluation, conducted in a challenging and realistic framework, suggest possible directions for future improvements in these smart sensors that are thoroughly discussed in the paper. Full article