Search results

Chatbots are increasingly engaged in service marketing. Some academics and managers think using anthropomorphism chatbots will improve positive attitudes and behaviors in the customer journey, but at a high degree of anthropomorphism, consumers may experience negative emotions such as fear and disgust due to the feeling that the robots resemble humans too much, which is known as the uncanny valley effect. Therefore, the authors aim to explore whether chatbot anthropomorphism will promote or limit the development of the customer journey and explore the moderating factors and the antecedent factors affecting consumers' perceptions of chatbot anthropomorphism.

The authors collected 72,782 unique data points from 42 articles and 82 samples using a meta-analysis. Based on the stimuli-organism-response (SOR) model, the impact of anthropomorphic chatbots on the consumer journey was discussed.

The authors’ findings show that chatbot anthropomorphism positively impacts the customer journey but not their negative attitudes. Further moderator analysis reveals that the impact depends on service result, chatbot gender and sample source. The chatbot anthropomorphism is significantly influenced by social presence cues, emotional message cues and mixed cues.

This research contributes to the chatbot anthropomorphism literature and offers guidance for managers on whether and how to enhance chatbot anthropomorphism to facilitate the customer journey and improve service sustainability.

This paper aims to study numerically the influence of the applied laser energy density and the porosity of the powder bed on the thermal behavior of the melt and the resultant instability of the liquid track.

A three-dimensional model was proposed to predict local powder melting process. The model accounts for heat transfer, melting, solidification and evaporation in granular system at particle scale. The proposed model has been proved to be a good approach for the simulation of the laser melting process.

The results shows that the applied laser energy density has a significantly influence on the shape of the molten pool and the local thermal properties. The relative low or high input laser energy density has the main negative impact on the stability of the scan track. Decreasing the porosity of the powder bed lowers the heat dissipation in the downward direction, resulting in a shallower melt pool, whereas pushing results in improvement in liquid track quality.

The randomly packed powder bed is calculated using discrete element method. The powder particle information including particle size distribution and packing density is taken into account in placement of individual particles. The effect of volumetric shrinkage and evaporation is considered in numerical model.

With the increasing awareness of global warming and the important role of last mile distribution in logistics activities, the purpose of this paper is to build an environmental and effective last mile distribution model considering fuel consumption and greenhouse gas emission, vehicle capacity and two practical delivery service options: home delivery (HD) and pickup site service (PS). This paper calls the problem as the capacitated pollution-routing problem with pickup and delivery (CPRPPD). The goal is to find an optimal route to minimize operational and environmental costs, as well as a set of optimal speeds over each arc, while respecting capacity constraints of vehicles and pickup sites.

To solve this problem, this research proposes a two-phase heuristic algorithm by combining a hybrid ant colony optimization (HACO) in the first stage and a multiple population genetic algorithm in the second stage. First, the HACO is presented to find the minimal route solution and reduce distribution cost based on optimizing the speed over each arc.

To verify the proposed CPRPPD model and algorithm, a real-world instance is conducted. Comparing with the scenario including HD service only, the scenario including both HD and PS option is more economical, which indicates that the CPRPPD model is more efficient. Besides, the results of speed optimization are significantly better than before.

The developed CPRPPD model not only minimizes delivery time and reduces the total emission cost, but also helps logistics enterprises to establish a more complete distribution system and increases customer satisfaction. The model and algorithm of this paper provide optimal support for the actual distribution activities of logistics enterprises in low-carbon environment, and also provide reference for the government to formulate energy-saving and emission reduction policies.

This paper provides a great space for the improvement of carbon emissions in the last mile distribution. The results show that the distribution arrangement including HD and PS services in the last mile adopting speed optimization can significantly reduce the carbon emission. Additionally, an integrated real-world instance is applied in this paper to illustrate the validity of the model and the effectiveness of this method.

This paper aims to develop a dynamic two-sided stable matching method based on preference information of the matching objects in uncertain environments, so as to solve the matching problem of cloud manufacturing tasks and services with load balancing.

For dynamic two-sided matching, due to the complexity of social environment and the limitation of human cognition, hesitation and fuzziness always exist in the process of multi-criteria assessment. First, in order to obtain the accurate preference information of each matching object, uncertain linguistic variables, uncertain preference ordinal and incomplete complementary matrices are used to evaluate multi-criteria preference information. This process is undertaken by considering the probability of each possible matching pair. Second, the preference information at different times is integrated by using the time-series weight to obtain the comprehensive satisfaction degree matrices of the matching objects. Further, the load adjustment parameter is used to increase the satisfaction degree of the matching objects. Afterward, a dynamic two-sided stable matching optimization model is constructed by considering stable matching conditions. The model aims to maximize the satisfaction degree and minimizes the difference in the satisfaction degree of matching objects. The optimal stable matching results can be obtained by solving the optimization model. Finally, a numerical example and comparative analysis are presented to demonstrate the characteristics of the proposed method.

Uncertain linguistic variables, uncertain preference orders and incomplete complementary matrices are used to describe multi-criteria preference information of the matching objects in uncertain environments. A dynamic two-sided stable matching method is proposed, based on which a DTSMDM (dynamic two-sided matching decision-making) model of cloud manufacturing with load balancing can be constructed. The study proved that the authors can use the proposed method to obtain stable matching pairs and higher matching objective value through comparative analysis and the sensitivity analysis.

A new method for the two-sided matching decision-making problem of cloud manufacturing with load balancing is proposed in this paper, which allows the matching objects to elicit language evaluation under uncertain environment more flexibly to implement dynamic two-sided matching based on preference information at different times. This method is suitable for dealing with a variety of TSMDM (two-sided matching decision-making) problems.

The purpose of this paper is to propose a valid and quantitative measurement method of knowledge diffusion efficiency for the knowledge collaboration networks (KCNs).

This paper builds a weighted KCN model with the node and edge weights. Based on the weighted KCN, the factors of knowledge diffusion efficiency are proposed and analyzed. Then, the knowledge transfer effect between two nodes is proposed and measured by comprehensively integrating the above factors. Furthermore, the main metric of efficiency of knowledge diffusion is proposed by modifying Latora and Marchiori’s model of efficiency of network.

A case is studied to illustrate the applicability of the proposed weighted network model and the knowledge diffusion efficiency measurement method. The results show the methods proposed in this paper can better measure and analyze the knowledge diffusion efficiency of KCNs than the traditional un-weighted methods and the subjective evaluation methods.

The real KCNs are always weighted networks. The weighted model of KCN can better reflect the real networks than the un-weighted model. Based on the weighted networks, the measurement methods proposed in this paper can more efficiently and accurately measure and evaluate the knowledge diffusion efficiency than the traditional methods. This study can help researchers to better understand knowledge diffusion theoretically, and provide managers with a decision support for knowledge management in practice.

The purpose of this paper is to report on a three‐year PMI2 project for the British Council in 2008, one of seven to develop and strengthen partnerships with Chinese institutions in employability and entrepreneurship. Involving a partnership between Leeds Metropolitan University England and the Zhejiang University of Technology Hangzhou, China, the aim has been to analyse the Hangzhou and Zhejiang economies and examine current Chinese company requirements for skills and talent and their implications for teaching and learning and graduate supply. This was intended to strengthen the existing partnerships at a civic level between Leeds and Hangzhou and the successful MA in Trade and Finance run by the two universities.

The paper draws on preliminary interview studies in China of Hangzhou companies in different industrial sectors to analyse the skill and talent needs of such companies, their demands for graduate talent in particular and their views about the adequacy of the supply of that talent from local and national universities.

The paper clarifies the relationship between talent demand and supply in China, especially with regard to graduate talent, and presents an original analysis of the skill needs of the Hangzhou economy.

The paper suggests ways in which universities in Zhejiang and China generally could strengthen their engagement with businesses over talent demand and supply, and how they could develop courses and programmes that more effectively bridge the gap between universities and businesses.

Because of the multiple design elements and complicated relationship among design elements of complex products design, it is tough for designers to systematically and dynamically express and manage the complex products design process.

To solve these problems, a supernetwork model of complex products design is constructed and analyzed in this paper. First, the design elements (customer demands, design agents, product structures, design tasks and design resources) are identified and analyzed, then the sub-network of design elements are built. Based on this, a supernetwork model of complex products design is constructed with the analysis of the relationship among sub-networks. Second, some typical and physical characteristics (robustness, vulnerability, degree and betweenness) of the supernetwork were calculated to analyze the performance of supernetwork and the features of complex product design process.

The design process of a wind turbine is studied as a case to illustrate the approach in this paper. The supernetwork can provide more information about collaborative design process of wind turbine than traditional models. Moreover, it can help managers and designers to manage the collaborative design process and improve collaborative design efficiency of wind turbine.

The authors find a new method (complex network or supernetwork) to describe and analyze complex mechanical product design.

This study examines the key determinants of subscription intentions for ChatGPT Plus (paid version) in business settings, focusing on tasks such as system quality, information support, service quality, perceived intelligence, goal-congruent outcome and self-efficacy.

The study utilized a survey of office workers, analyzed through structural equation modeling, to explore these determinants.

The results demonstrate that system quality, service quality and perceived intelligence significantly influence satisfaction, while service quality and perceived intelligence also impact goal-congruent outcomes. Contrary to traditional models, satisfaction does not significantly correlate with usage. Instead, a significant relationship is observed between goal-congruent outcomes and usage. Self-efficacy emerges as a crucial predictor of subscription intentions, further underlined by the significant impact of usage on subscription intention.

The study’s focus on office workers and a single artificial intelligence (AI) chatbot type may limit generalizability. Its findings illuminate several avenues for future research, particularly in diversifying the context and demographics studied.

This research offers actionable insights for businesses and practitioners in the implementation of AI chatbots. It highlights the importance of enhancing system quality, personalization and user confidence to boost subscription intentions, thereby guiding strategies for user engagement and technology adoption.

This study pioneers in investigating subscription intentions towards AI chatbots, particularly ChatGPT, providing a novel framework that expands upon traditional user behavior theories.