Collaborative robots are increasingly deployed in Industry 5.0 disassembly cells, yet how operators adapt across repeated cobot interactions is rarely characterised multimodally. The Day-2 axis of the MultiPhysio-HRC corpus was analysed: forty-two participants performed up to five repetitions each of a Fanuc CRX-20 cobot-assisted disassembly and a matched manual control, with 12-channel dry EEG, ECG, EDA, EMG and respiration recorded throughout; primary inference was based on the paired-analysis cohort (n = 39, ≥ 3 repetitions per context). Mixed-effects models and per-subject physiological regressions revealed a within-subject adaptation signature confined to cobot: state anxiety, workload, frustration and arousal declined while dominance grew, whereas manual produced flat or worsening trajectories. Repeated-measures analyses showed stronger aggregate physiological-state coupling during cobot than manual work (subject-level permutation p ≈ .001), although no individual channel showed a reliable context-specific effect. Cross-subject prediction of individual adaptation slopes was largely unsuccessful.

As home appliance interfaces become increasingly digitalized, users with visual impairment or spinal cord injury (SCI) face growing barriers to operating essential household devices. This study examines two prototype designs intended to improve accessibility: an acrylic panel with engraved grooves and Braille markings for washing machine touch panels, and a swing button that transforms pushing into a top-down pressing motion for microwave doors. Nine participants (five with visual impairment; four with SCI) evaluated the prototypes using a study-specific questionnaire combining Likert-scale and open-ended items. Both prototypes improved tactile guidance, reduced unintended activations, and enhanced usability within reachable ranges; an unexpected benefit was reduced glare for wheelchair users. However, tactile cues alone were insufficient for visually impaired users, who emphasized the need for voice feedback. These preliminary findings, though limited by sample size, support integrating accessibility into mainstream appliances rather than relegating it to specific products.

Efficient waste classification is crucial for promoting recycling and achieving sustainable waste management. Real-world waste streams, however, often include mixed, deformed, and contaminated items, making manual sorting inefficient and error prone. A deep learning-based system for multi-class classification of heterogeneous waste using the RealWaste dataset is presented in this paper, which reflects actual disposal conditions such as cluttered backgrounds and overlapping materials. We fine-tune and evaluate several convolutional neural networks (CNNs), including InceptionV3, ResNet101, DenseNet, VGG, EfficientNet, and MobileNet. Among these, ResNet101 demonstrated the best performance, achieving a validation accuracy of 98.86%, loss of 0.0379, and 0.99 as F1 score. We also introduce hybrid models (e.g., ResNet101 + InceptionV3), which improved precision in complex categories such as textiles and miscellaneous trash. Furthermore, a confidence score evaluation strategy is proposed to assess model reliability, revealing high confidence (≥ 0.95) for visually distinct classes like vegetation, plastic, and food organics. Our findings establish a robust and scalable benchmark for deploying intelligent waste classification systems in real-world, sustainability-driven environments.

The movement of visually impaired people is still limited, and they often require assistance from others. In this study, along with the development of autonomous driving technology, a future mobility design that will help visually impaired people conveniently move around was proposed. The Double-Diamond model, a representative UX evaluation method, was revised and used for the evaluation. After discovering the mobility problems of the visually impaired, we developed the problem into an idea and designed future mobility based on the idea. Then, it was delivered to visually impaired people, and a utility test was performed on the new concept and functions. Six functions were proposed in scenarios for each moving process, and the evaluation results showed that drop-off notification using multi-senses showed the highest utilization. It is hoped that the expansion of self-driving vehicles will increase the mobility of visually impaired people with difficulty driving.

Purpose Breast cancer encompasses various subtypes with distinct prognoses, necessitating accurate stratification methods. Current techniques rely on quantifying gene expression in limited subsets. Given the complexity of breast tissues, effective detection and classification of breast cancer is crucial in medical imaging. This study introduces a novel method, MPa-DCAE, which uses a multi-patch-based deep convolutional auto-encoder (DCAE) framework combined with VGG19 to detect and classify breast cancer in histopathology images. Methods The proposed MPa-DCAE model leverages the hierarchical feature extraction capabilities of VGG19 within a DCAE framework, designed to capture intricate patterns in histopathology images. By using a multi-patch approach, regions of interest are extracted from pathology images to facilitate localized feature learning, enhancing the model's discriminatory power. The auto-encoder component enables unsupervised feature learning, increasing resilience and adaptability to variations in image features. Experiments were conducted at various magnifications on the CBIS-DDSM and MIAS datasets to validate model performance. Results Experimental results demonstrated that the MPa-DCAE model outperformed existing methods. For the CBIS-DDSM dataset, the model achieved a precision of 97.96%, a recall of 94.85%, and an accuracy of 98.36%. For the MIAS dataset, it achieved a precision of 97.99%, a recall of 97.2%, and an accuracy of 98.95%. These results highlight the model's robustness and potential for clinical application in computer-assisted diagnosis. Conclusion The MPa-DCAE model, integrating VGG19 and DCAE, proves to be an effective, automated approach for diagnosing breast cancer. Its high accuracy and generalizability make it a promising tool for clinical practice, potentially improving patient care in histopathology-based breast cancer diagnosis.

The most used self-assessment method for assessing driving style is the Multidimensional Driving Style Inventory (MDSI). This study aims to adapt the MDSI to Korean drivers (MDSI-K) and confirm the eight-factor driving style structure of the original version. Six hundred forty drivers aged 20-70 agreed to participate in this study. All participants had at least one year of driving experience. Confirmatory factor analysis was performed to determine whether the original MDSI fit the Korean driving style structure, and the goodness-of-fit values were not acceptable. Through correlation analysis, 11 survey questions were removed, and MDSI-K, which has the same structure as the original MDSI, was derived. Subsequent exploratory factor analysis yielded an 8-factor structure with several cross-loadings between factors. For more rigorous validation, Exploratory Structural Equation Modeling was employed, resulting in a refined model comprising 33 items across 8 factors, which incorporated these cross-loadings in the MDSI-K construction. We propose that these cross-loadings represent cultural differences in driving styles. In conclusion, the results of this study show that there is a common concept of driving style that has been previously investigated in various countries. This can be seen as a similarity in the driving culture commonly used across countries. On the other hand, the results of the different structures or survey questions for the same questionnaire by country and culture show differences between cultures.

The rapid growth of the electric scooter (e-Scooter) market and its increasing urban use necessitate research into rider behavioral patterns and their impact on traffic safety. This study aims to design and validate an e-scooter rider behavior questionnaire (eSRBQ) to address this need. The eSRBQ was validated based on established tools like the Driver Behavior Questionnaire (DBQ) and the Cyclist Behavior Questionnaire (CBQ). A sample of 445 participants completed the eSRBQ, which comprised three factors: violations, errors, and positive behaviors. Confirmatory factor analysis (CFA) validated the questionnaire’s structure. The eSRBQ demonstrated good fit indices (Chi-squared/ df = 3.651, SRMR = 0.047, RMSEA = 0.077, CFI = 0.941, TLI = 0.932) and high internal consistency (Cronbach’s alpha > .7 for all factors). The study found significant differences in violations, errors, and positive behaviors based on collision history. Age and riding frequency also influenced behavior, with younger and less frequent riders displaying more violations. These findings highlight the necessity for targeted safety interventions and continuous monitoring of e-scooter rider behavior as their use expands. This study contributes to urban transportation safety by providing a validated tool for assessing e-scooter rider behavior and identifying critical factors for improving road safety.

In tree-based algorithms like random forest and deep forest, due to the presence of numerous inefficient trees and forests in the model, the computational load increases and the efficiency decreases. To address this issue, in the present paper, a model called Automatic Deep Forest Shrinkage (ADeFS) is proposed based on shrinkage techniques. The purpose of this model is to reduce the number of trees, enhance the efficiency of the gcforest, and reduce computational load. The proposed model comprises four steps. The first step is multi-grained scanning, which carries out a sliding window strategy to scan the input data and extract the relations between features. The second step is cascade forest, which is structured layer-by-layer with a number of forests consisting of random forest (RF) and completely random forest (CRF) within each layer. In the third step, which is the innovation of this paper, shrinkage techniques such as LASSO and elastic net (EN) are employed to decrease the number of trees in the last layer of the previous step, thereby decreasing the computational load, and improving the gcforest performance. Among several shrinkage techniques, elastic net (EN) provides better performance. Finally, in the last step, the simple average ensemble method is employed to combine the remaining trees. The proposed model is evaluated by Monte Carlo simulation and three real datasets. Findings demonstrate the superior performance of the proposed ADeFS-EN model over both gcforest and RF, as well as the combination of RF with shrinkage techniques.

Ensembling is a powerful technique to obtain the most accurate results. In some cases, the large number of learners in ensemble learning mostly increases both computational load during the test phase and error rate. To solve this problem, in this paper we propose an Ensemble of Reduced Deep Regression (ERDeR) model, which is a combination of Deep Regressions (DRs), shrinkage methods, and ensemble approaches. The framework of the proposed model contains three phases. The first phase includes base regressions in which parallel DRs are used as learners. The role of these DRs is to extract features of input data and make prediction. In the second phase, to automatically reduce and select the most suitable DRs, shrinkage methods such as Least Absolute Shrinkage and Selection Operator (LASSO) and Elastic Net (EN) are employed. These models are compared with the non-shrinkage model. The last phase is ensemble phase, which consists of three different ensemble methods namely Multi-Layer Perceptron (MLP), Weighted Average (WA), and Simple Average (SA). These ensemble methods are used to aggregate the remaining learners from previous steps. Finally, the proposed model is applied to Monte Carlo simulation data and three real datasets including Boston House Price, Real Estate Valuation and Gold Price per Ounce. The results show that after applying the shrinkage methods the error rate is significantly reduced and the model accuracy is increased. Accordingly, the results of combining shrinkage methods and ensemble approaches not only decreased the computational load during test phase, but also increased the model accuracy.

The purpose of this study is to examine the effect of information quality(IQ) on customer satisfaction and loyalty within cross-buying contexts in the telecommunications industry. The research uses a survey of 215 participants and a structural equation model to analyze how IQ influences loyalty, employing Cronbach's alpha(CA), correlation, and confirmatory factor analysis(CFA) for measurement validity. Results indicate that all indicators, including word of mouth(WOM) and accessibility IQ, show high reliability(CA > 0.7). The model demonstrates that customer satisfaction—mediated by contextual IQ—affects loyalty significantly. This investigation highlights the critical role of IQ in fostering loyalty and satisfaction in Korea's competitive telecom market, emphasizing the need for precise information delivery while avoiding spam. It also points to the necessity of future research on user needs in loyalty building, particularly in scenarios where secondary purchases accompany primary ones.

As the population of electric vehicles (EVs) continues to grow, managing and enhancing the Electric Vehicle Charging Experience (eCX) has become an inevitable challenge. However, the research community has not given comprehensive attention to the eCX, often neglecting interactions between key components: EVs, chargers, and mobile apps. This research addresses the gap by identifying heuristic evaluation criteria to measure the problems of existing EV chargers for effective management. The charging process was analyzed hierarchically to establish criteria, and its relevance to previously established heuristics was evaluated. Domain experts then verified 27 criteria in 7 tasks, whether these can effectively identify issues affecting eCX through ratings and discussions. As a result of quantitative analysis of the ratings and qualitative examination of the discussions, formulated criteria can offer comprehensive and valuable insights spanning various eCX components and tasks. This research can guide designers in enhancing the eCX for current and prospective users.

This study aims to investigate the recent literature on auditory experiences within automotive environments and discusses potential future research directions in this area. Forty-six papers obtained through the PRISMA protocol were selected from literature published over the past 15 years. The collected literature was categorized based on engine type, and a comparative analysis of research trends in the automotive industry was conducted, explicitly focusing on internal combustion vehicles (ICVs) and electric vehicles (EVs). A network analysis was performed utilizing the keywords of the papers to identify the predominant research topics. The analysis revealed research topics actively studied in existing ICV research but not covered in EV and newly emerging research topics in the EV field. The study proposes future research topics related to auditory experience design. It aims to provide insight into the design of auditory experiences in automobiles, particularly as the automotive paradigm expands to include electric and autonomous vehicles.

Label sparsity in multivariate time series (MTS) makes using label information for practical applications challenging. Thus, unsupervised representation learning methods have gained attention to learn effective representations suitable for various MTS tasks without relying on labels. Recently, contrastive learning has emerged as a promising approach to generate robust representations by capturing underlying MTS information. However, the existing methods have some limitations, such as insufficient consideration of cross-variable relationships of MTS and high sensitivity to positive pairs. Therefore, we proposed a novel spatio-temporal contrastive representation learning method (STCR) designed to address these limitations. STCR focuses on learning robust representations by encouraging spatio-temporal consistency, which comprehensively considers spatial information as well as temporal dependencies in MTS. The results of extensive experiments on MTS classification and forecasting tasks demonstrate the efficacy of STCR in generating high-quality representations, achieving state-of-the-art performance on both tasks.

PurposeThis study proposes a therblig-based evaluation technique as a new accessibility tool for physical products like home appliances that spinal cord injured users occasionally use.Material and MethodsThis study recruited nine spinal cord injured users for the interview and observation regarding home appliance usage and analytically structured their usage behaviors using therbligs. The therblig notations eventually referred to actual and potential accessibility issues that spinal cord injured users would encounter when using the home appliances.ResultsThe primary therblig operations causing accessibility issues for spinal cord injured users were 'reach,' 'move,' 'grasp,' 'position,' and 'use', corresponding to their disability characteristics. In addition, this study proposed a new effective therblig called "hook," which is suitable for better representation of user behavior and accessibility evaluation of spinal cord users.ConclusionThis study provided an interaction-based accessibility evaluation technique, which is easy to learn and apply, especially for physical products.

Ion homeostasis, which is regulated by ion channels, is crucial for intracellular signaling. These channels are involved in diverse signaling pathways, including cell proliferation, migration, and intracellular calcium dynamics. Consequently, ion channel dysfunction can lead to various diseases. In addition, these channels are present in the plasma membrane and intracellular organelles. However, our understanding of the function of intracellular organellar ion channels is limited. Recent advancements in electrophysiological techniques have enabled us to record ion channels within intracellular organelles and thus learn more about their functions. Autophagy is a vital process of intracellular protein degradation that facilitates the breakdown of aged, unnecessary, and harmful proteins into their amino acid residues. Lysosomes, which were previously considered protein-degrading garbage boxes, are now recognized as crucial intracellular sensors that play significant roles in normal signaling and disease pathogenesis. Lysosomes participate in various processes, including digestion, recycling, exocytosis, calcium signaling, nutrient sensing, and wound repair, highlighting the importance of ion channels in these signaling pathways. This review focuses on different lysosomal ion channels, including those associated with diseases, and provides insights into their cellular functions. By summarizing the existing knowledge and literature, this review emphasizes the need for further research in this field. Ultimately, this study aims to provide novel perspectives on the regulation of lysosomal ion channels and the significance of ion-associated signaling in intracellular functions to develop innovative therapeutic targets for rare and lysosomal storage diseases.

As the technology gap decreases and competition intensifies, information quality (IQ) has become one of the most important factors when customers buy products. Despite the overflowing-information environment, few studies have defined the properties of information quality and their impacts on customer purchase decisions. Therefore, this study examined customers’ purchase behaviors with IQ attributes in different contexts of cross-buying and repurchasing. After surveying 150 customers of Korean telecommunication companies, constructs and their effects were evaluated from factor analysis and regression analysis, respectively. This study revealed that different attributes of IQ were effective for different contexts of customer purchases. Thus, different information attributes should be strategically emphasized for customers under different contexts of purchase decisions.

Electronic word of mouth (e-WOM) influences consumer decision-making. Since consumers' affective experiences for products are vast, research is needed to understand and categorize them accurately. In this paper, we developed a deep learning-based clustering algorithm for categorizing consumer sentiment in product reviews and explored the applicability of this algorithm. A Deep Attentive Self-Organizing Map (DASOM) was created by noting individualized sentimental characteristics of each review and interpreting why each review was included in a particular cluster. As a result of analyzing 4941 reviews of Amazon, one of online commerce platforms, it was confirmed that sentiment classification through DASOM could be effectively used to categorize implicit affective experiences of consumers. DASOM was effective in identifying the relationship between multi-dimensional affective elements that were difficult to derive from TF-IDF. Using the proposed methodology, it is expected to provide practical information for companies that design products considering consumer affection.

In e-commerce, customer feedback has become an essential source of insight into a product or service's user experience (UX). The study of UX helps to integrate customers' potential needs into the product's design. Because customer reviews in e-commerce are not structured and categorized, it is necessary to analyze UX based on customer opinions systematically. This study tries to structure UX in a product's positive/negative context through a neural network-based self-organizing map (SOM). As a result of analyzing 10,482 reviews on wireless earbuds in BestBuy, an electronic product e-commerce platform, it was confirmed that it is a suitable method for categorizing user experiences between reviews and deriving important factors. In particular, the difference in core UX elements by positive/negative context of the product was verified based on the star rating. The results of this study are expected to contribute to product improvement and business improvement that reflect customer needs by companies or designers who design products for end-users.

This paper investigates the ChatGPT research landscape, key themes, and connections of recent studies in three major domains of computer science (CS), social science (SS), and medical (MED). 822 studies containing ‘ChatGPT’ in the title, keywords, and abstract were collected from research databases such as Scopus, Web of Science, and ScienceDirect. Excluding duplicates and papers without keywords, the distribution across domains for the remaining 660 papers is as follows: 189 in CS, 238 in SS, and 395 in MED.Using semantic network analysis, we examine keyword connections and concepts shared between papers, to understand how ChatGPT is being utilized across various fields. We perform modularity analysis to identify clusters of related studies and emerging research trends. The results of this study enhances the understanding of in three major domains. Our insights serve as a valuable guide for creating customized ChatGPT services, and help researchers make informed decisions for advancing the AI's applications.

This study investigates optimal predictive algorithms to discern human emotional responses provoked by vehicular engine acceleration sounds. It defines two affective attributes, "sporty" and "luxury," linked to four psychoacoustic parameters. The study employs linear regression, multi-layer perceptron artificial neural network, and support vector regression to compute root mean square errors for these attributes. To mitigate overfitting, a 10-fold cross-validation technique is utilized. Results reveal the potential of a psychoacoustic-based prediction algorithm for engine sound affectivity. Artificial neural networks have emerged as the superior choice for forecasting human emotional responses to engine sounds. The study's implications lie in its potential to predict consumer satisfaction with engine sounds, offering a competitive advantage in the market by leveraging the analysis of affective dimensions.