Search Results (93)

Production involves processes such as raw material extraction, energy consumption, and waste management, which can lead to significant environmental consequences. Therefore, supplier selection based not only on technical performance but also on environmental sustainability criteria has become a fundamental component of eco-friendly manufacturing strategies. Moreover, in the selection of electric vehicle batteries, it is essential to consider customer demands alongside environmental factors. Accordingly, selected suppliers should fulfill company expectations while also reflecting the “voice” of the customer. The objective of this study is to propose an integrated approach for green supplier selection by taking into account various environmental performance requirements and criteria. The proposed approach evaluates battery suppliers with respect to both customer requirements and green criteria. To construct the relational structure, the DEMATEL method was employed to analyze the interrelationships among customer requirements (CRs). Subsequently, the Quality Function Deployment (QFD) model was used to establish a central relational matrix that captures the degree of correlation between each pair of supplier selection criteria and CRs. Finally, to evaluate and rank alternative suppliers, the Interval Type-2 Fuzzy VIKOR (IT2 F-VIKOR) method was applied. The hybrid approach proposed by us, integrating DEMATEL, QFD, and IT2 F-VIKOR, offers significant improvements over traditional methods. Unlike previous approaches that focus independently on customer preferences or supplier criteria, our model provides a unified evaluation by considering both dimensions simultaneously. Furthermore, the use of Interval Type-2 Fuzzy Logic enables the model to better manage uncertainty and ambiguity in expert judgments, yielding more reliable results compared to conventional fuzzy approaches. Additionally, the applicability of the model has been demonstrated through a real-world case study, confirming its practical relevance and robustness in the selection of green suppliers for electric vehicle battery procurement. Full article

Electric bicycles in elevators pose serious safety hazards. Fires in the confined space make escape difficult, and recent accidents involving e-bike fires have caused casualties and property damage. To prevent e-bikes from entering elevators and improve public safety, this design employs the Nezha development board as the upper computer for visual detection. It uses deep learning algorithms to recognize hazards like e-bikes. The lower computer orchestrates elevator controls, including voice alarms, door locking, and emergency halt. The system comprises two parts: the upper computer uses the YOLOv11 model for target detection, trained on a custom e-bike image dataset. The lower computer features an elevator control circuit for coordination. The workflow covers target detection algorithm application, dataset creation, and system validation. The experiments show that the YOLOv11 demonstrates superior e-bike detection performance, achieving 96.0% detection accuracy and 92.61% mAP@0.5, outperforming YOLOv3 by 6.77% and YOLOv8 by 15.91% in mAP, significantly outperforming YOLOv3 and YOLOv8. The system accurately identifies e-bikes and triggers safety measures with good practical effectiveness, substantially enhancing elevator safety. Full article

Parkinson’s disease (PD) is a progressive neurodegenerative disorder leading to movement impairment, cognitive decline, and psychiatric symptoms. Key manifestations of PD include bradykinesia (the slowness of movement), changes in voice or speech, and gait disturbances. The quantification of neurological disorders through voice analysis has emerged as a rapidly expanding research domain, offering the potential for non-invasive and large-scale monitoring. This review explores existing research on the application of machine learning (ML) in speech, voice, and language processing for the diagnosis of PD. It comprehensively analyzes current methodologies, highlights key findings and their associated limitations, and proposes strategies to address existing challenges. A systematic review was conducted following PRISMA guidelines. We searched four databases: PubMed, Web of Science, Scopus, and IEEE Xplore. The primary focus was on the diagnosis, detection, or identification of PD through voice, speech, and language characteristics. We included 34 studies that used ML techniques to detect or classify PD based on vocal features. The most used approaches involved free speech and reading-speech tasks. In addition to widely used feature extraction toolkits, several studies implemented custom-built feature sets. Although nearly all studies reported high classification performance, significant limitations were identified, including challenges in comparability and incomplete integration with clinical applications. Emerging trends in this field include the collection of real-world, everyday speech data to facilitate longitudinal tracking and capture participants’ natural behaviors. Another promising direction involves the incorporation of additional modalities alongside voice analysis, which may enhance both analytical performance and clinical applicability. Further research is required to determine optimal methodologies for leveraging speech and voice changes as early biomarkers of PD, thereby enhancing early detection and informing clinical intervention strategies. Full article

This paper tackles the logistics dilemma of how to meet customer expectations while at the same time respecting the internal processes and financial interests of the company and ensuring long-term sustainability. In this paper, integrated Quality Function Deployment (QFD) and Balanced Scorecard (BSC) techniques developed a method for the structured planning of logistics strategies. BSC, combined with QFD, gives the opportunity not only to “translate” the voice of the customer but also to focus on the company’s interests from four perspectives. For example, for products, we evaluated the interactions between different expectations, and the focus was on the disputes that arise during the expectations. The result of this paper is that Extended QFD provides a new method to formulate the various requirements. This method is suitable for creating a sustainable logistics strategy. Full article

The adoption of Intelligent Virtual Assistants (IVAs) in the banking sector presents new opportunities to enhance customer service efficiency, reduce operational costs, and modernize service delivery channels. However, the factors driving IVA adoption and usage, particularly in specific regional contexts such as Portugal, remain underexplored. This study examined the determinants of IVA adoption intention and actual usage in the Portuguese banking sector, drawing on the Technology Acceptance Model (TAM) as its theoretical foundation. Data were collected through an online questionnaire distributed to 154 banking customers after they interacted with a commercial bank’s IVA. The analysis was conducted using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings revealed that perceived usefulness significantly influences the intention to adopt, which in turn significantly impacts actual usage. In contrast, other variables—including trust, ease of use, anthropomorphism, awareness, service quality, and gendered voice—did not show a significant effect. These results suggest that Portuguese users adopt IVAs based primarily on functional utility, highlighting the importance of outcome-oriented design and communication strategies. This study contributes to the understanding of technology adoption in mature digital markets and offers practical guidance for banks seeking to enhance the perceived value of their virtual assistants. Full article

Postal services have re-emerged across numerous emerging economies worldwide as essential logistics providers, harnessing their vast coverage and dependability in the face of expanding e-commerce platforms and technological innovations. This study investigates India Post, one of the largest postal networks globally, to determine the key logistics service parameters prioritized by customers in southern India. Quantitative data obtained from 255 India Post end-users were evaluated using the logistics service quality (LSQ) scale, assessing eight dimensions including information quality, timeliness, ordering procedure, order accuracy, order condition, personal contact quality, order discrepancy handling, and order release quantities. The Analytical Hierarchy Process (AHP) ranked these elements, while Quality Function Deployment (QFD) bridged customer expectations with service improvements. The findings highlight the need to improve sorting and distribution processes to meet customer demands for timely, high-quality delivery. By refining logistics efficiency, this study provides suggestions and recommendations for boosting satisfaction and profitability, shedding light on service-led economic advancement for postal services in emerging economies. These insights equip postal service providers to cultivate loyalty and maintain competitiveness within the dynamic logistics landscape. Full article

Moroccan Law 55.19 aims to streamline administrative procedures, fostering trust between citizens and public administrations. To implement this law effectively and enhance public service quality, it is essential to use the Moroccan dialect to involve a wide range of people by leveraging Natural Language Processing (NLP) techniques customized to its specific linguistic characteristics. It is worth noting that the Moroccan dialect presents a unique linguistic landscape, marked by the coexistence of multiple texts. Though it has emerged as the preferred medium of communication on social media, reaching wide audiences, its perceived difficulty of comprehension remains unaddressed. This article introduces a new approach to addressing these challenges. First, we compiled and processed a dataset of Moroccan dialect requests for public administration documents, employing a new augmentation technique to enhance its size and diversity. Second, we conducted text classification experiments using various machine learning algorithms, ranging from traditional methods to advanced large language models (LLMs), to categorize the requests into three classes. The results indicate promising outcomes, with an accuracy of more than 80% for LLMs. Finally, we propose a chatbot system architecture for deploying the most efficient classification algorithm. This solution also contains a voice assistant system that can contribute to the social inclusion of illiterate people. The article concludes by outlining potential avenues for future research. Full article

With advances in digital transformation (DX) in education and digital technologies becoming more deeply integrated into educational settings, global demand for video-based learning materials continues to rise, resulting in substantial effort being required from teachers to create e-learning videos. Furthermore, while many existing services offer visual content, they primarily rely on templates, making it challenging to design custom content that addresses specific needs. In this study, we develop a web service that facilitates e-learning video creation through integrated artificial intelligence (AI) and digital human technology. This service enhances educational content by integrating digital human characters and voice synthesis technologies, aiming to create comprehensive e-learning videos by incorporating visual motion and synchronized audio into educational content. In addition, this service also aims to enable the creation of engaging content through advanced visuals and animations, effectively maintaining learner interest. Full article

This paper proposes a solution for guiding visually impaired people to reach predefined locations marked with preregistered passive ultra-high-frequency RFID tags inside public buildings (e.g., secretary’s offices and information desks). Our approach employs an unmanned ground vehicle guidance system that assists customers in following predefined routes. The solution also includes a methodology for recording the best routes between all possible locations that may be visited. When reaching the destination, the system will read the tag, extract all the associated information from a database, and translate it into an audio format played into the user’s headphones. The system includes functionalities such as recording and playback of prerecorded routes, voice commands, and audio instructions. By describing the software and hardware architecture of the proposed guiding systems prototype, we show how combining ultra-high-frequency RFID technology with unmanned ground vehicle guiding systems equipped with ultrasonic, grayscale, hall sensors, and voice interfaces allows the development of accessible, low-cost guiding systems with increased functionalities. Moreover, we compare and analyze two different modes of route recording based on line following and manual recording, obtaining a performance regarding route playback with deviations under 10% for several basic scenarios. Full article

Internet platforms and self-media have become vital online communities for promoting positive reputations for hotels. Previous studies have primarily focused on enhancing positive electronic word-of-mouth (eWOM) through improvements in hotel infrastructure and staff services. As hotels deepen their digital transformation, the application of various artificial intelligence technologies in hotel service encounters significantly impacts the service experience. This study explores the effects of voice assistant (VA) attributes on the online reputation of hotels. Specifically, it examines how the attributes of VAs (anytime connectivity, information association, and interactivity) influence positive customer evaluations in hotels. Utilizing a questionnaire survey method, we collected 529 valid questionnaires offline and employed structural equation modeling along with the PROCESS plugin in SPSS to conduct path analysis, as well as mediation and moderation effect analyses. The results indicate that perceived value and the existence of human–AI rapport mediate the impact of VA attributes on positive eWOM, although the direct effect of some attributes (information association) was not supported. Furthermore, anytime connectivity enhances the influence on human–AI rapport through social presence, while privacy concerns negatively affect the relationship between perceived value and intentions to engage in eWOM. These insights are critical for hotels seeking to maximize the benefits of digital transformation. Full article

Effective corporate social responsibility (CSR) implementation is essential for construction enterprises to achieve sustainable development. However, existing reviews on CSR indicators and performance measures predominantly employ a single review method or focus on non-construction sectors, with limited exploration of their interrelationships. To address this gap, this state-of-the-art review synthesizes findings from 77 relevant papers published over the past two decades in Scopus, adopting a combined methodological approach that integrates science mapping and systematic review techniques. The scientometric analysis, conducted using VOSviewer, examines annual publication trends, key journals, prominent keywords, contributing countries, and influential documents. A subsequent systematic discussion utilizing content analysis identifies seven critical CSR indicators (e.g., environmental sustainability, corporate practices, and employee well-being) and eight performance dimensions (e.g., customer satisfaction and corporate reputation). A conceptual linkage framework is developed to elucidate the relationships between these indicators and performance dimensions, highlighting the most influential CSR factors. To enhance the robustness of the findings, a post-survey interview method is employed to validate and compare the systematic discussion results, revealing several cognitive gaps between academic perspectives and industry practices. Finally, future research directions and study limitations are discussed. By integrating the mixed-review results with voices of the construction industry, this review provides an objective and holistic reference for CSR scholars in the construction sector and offers managerial and policy insights for industry stakeholders and policymakers. Full article

Academic publications from the Web of Science Core Collection on “e-business” and “artificial intelligence” (AI) are investigated to reveal the role of AI, extract latent themes and identify potential research topics. The proposed methodology includes relevant graphical representations (trends, co-occurrence networks, Sankey diagrams), sentiment analyses and latent topics identification. A renewed interest in these publications is evident post-2018, with a sharp increase in publications around 2020 that can be attributed to the COVID-19 pandemic. Chinese institutions dominate the collaboration network in e-business and AI. Keywords such as “business transformation”, “business value” and “e-business strategy” are prominent, contributing significantly to areas like “Operations Research & Management Science”. Additionally, the keyword “e-agribusiness” recently appears connected to “Environmental Sciences & Ecology”, indicating the application of e-business principles in sustainable practices. Although three sentiment analysis methods broadly agree on key trends, such as the rise in positive sentiment over time and the dominance of neutral sentiment, they differ in detail and focus. Custom analysis reveals more pronounced fluctuations, whereas VADER and TextBlob present steadier and more subdued patterns. Four well-balanced topics are identified with a coherence score of 0.66 using Latent Dirichlet Allocation, which is a probabilistic generative model designed to uncover hidden topics in large text corpora: Topic 1 (29.8%) highlights data-driven decision-making in e-business, focusing on AI, information sharing and technology-enabled business processes. Topic 2 (28.1%) explores AI and Machine Learning (ML) in web-based business, emphasizing customer service, innovation and workflow optimization. Topic 3 (23.6%) focuses on analytical methods for decision-making, using data modeling to enhance strategies, processes and sustainability. Topic 4 (18.5%) examines the semantic web, leveraging ontologies and knowledge systems to improve intelligent systems and web platforms. New pathways such as voice assistance, augmented reality and dynamic marketplaces could further enhance e-business strategies. Full article

Complaint processing is of great importance for companies because it allows them to understand customer satisfaction levels, which is crucial for business success. It allows them to show the real perceptions of users and thus visualize the problems, which are regularly processed from oral or written natural language, derived from the provision of a service. In addition, the treatment of complaints is relevant because according to the laws of each country, companies have the obligation to respond to these complaints in a specified time. The specialized literature mentions that enterprises lost USD 75 billion due to poor customer service, highlighting that companies need to know and understand customer perceptions, especially emotions, and product reviews to gain insight and learn about customer feedback because of the importance of the voice of the customer for an organization. In general, it is evident that there is a need for research related to computational language processing to handle user requests. The authors show great interest in computational techniques for the processing of this information in natural language and how this could contribute to the improvement of processes within the productive sector. This work searches in indexed journals for information related to computational methods for processing relevant data from user complaints. It is proposed to apply a systematic literature review (SLR) method combining literature review guides by Kitchenham and the PRISMA statement. The systematic process allows the extraction of consistent information, and after applying it, 27 articles were obtained from which the analysis was conducted. The results show various proposals using linguistic, statistical, machine learning, and hybrid methods. We find that most authors combine Natural Language Processing (NLP) and Machine Learning (ML) to create hybrid methods. The methods extract relevant information from complaints of the customers in natural language in various domains, such as government, medical, banks, e-commerce, public services, agriculture, customer service, environmental, and tourism, among others. This work contributes as support for the creation of new systems that can give companies a significant competitive advantage due to their ability to reduce the response time of the complaints as established by law. Full article

Patient-level grouped data are prevalent in public health and medical fields, and multiple instance learning (MIL) offers a framework to address the challenges associated with this type of data structure. This study compares four data aggregation methods designed to tackle the grouped structure in classification tasks: post-mean, post-max, post-min, and pre-mean aggregation. We developed a customized AI pipeline that incorporates twelve machine learning algorithms along with the four aggregation methods to detect Parkinson’s disease (PD) using multiple voice recordings from individuals available in the UCI Machine Learning Repository, which includes 756 voice recordings from 188 PD patients and 64 healthy individuals. Seven performance metrics—accuracy, precision, sensitivity, specificity, F1 score, AUC, and MCC—were utilized for model evaluation. Various techniques, such as Bag Over-Sampling (BOS), cross-validation, and grid search, were implemented to enhance classification performance. Among the four aggregation methods, post-mean aggregation combined with XGBoost achieved the highest accuracy (0.880), F1 score (0.922), and MCC (0.672). Furthermore, we identified potential trends in selecting aggregation methods that are suitable for imbalanced data, particularly based on their differences in sensitivity and specificity. These findings provide meaningful implications for the further exploration of grouped imbalanced data. Full article

This paper proposes using a web crawler to organize website content as a dialogue tree in some domains. We build an intelligent customer service agent based on this dialogue tree for general usage. The encoder-decoder architecture Seq2Seq is used to understand natural language and then modified as a bi-directional LSTM to increase the accuracy of the polysemy cases. The attention mechanism is added in the decoder to improve the problem of accuracy decreasing as the sentence grows in length. We conducted four experiments. The first is an ablation experiment demonstrating that the Seq2Seq + Bi-directional LSTM + Attention mechanism is superior to LSTM, Seq2Seq, Seq2Seq + Attention mechanism in natural language processing. Using an open-source Chinese corpus for testing, the accuracy was 82.1%, 63.4%, 69.2%, and 76.1%, respectively. The second experiment uses knowledge of the target domain to ask questions. Five thousand data from Taiwan Water Supply Company were used as the target training data, and a thousand questions that differed from the training data but related to water were used for testing. The accuracy of RasaNLU and this study were 86.4% and 87.1%, respectively. The third experiment uses knowledge from non-target domains to ask questions and compares answers from RasaNLU with the proposed neural network model. Five thousand questions were extracted as the training data, including chat databases from eight public sources such as Weibo, Tieba, Douban, and other well-known social networking sites in mainland China and PTT in Taiwan. Then, 1000 questions from the same corpus that differed from the training data for testing were extracted. The accuracy of this study was 83.2%, which is far better than RasaNLU. It is confirmed that the proposed model is more accurate in the general field. The last experiment compares this study with voice assistants like Xiao Ai, Google Assistant, Siri, and Samsung Bixby. Although this study cannot answer vague questions accurately, it is more accurate in the trained application fields. Full article