Abstract
Chatbots are website artificial intelligence-based and automated customer support tools to improve the customer experience, to reduce costs, and to improve service quality. This study aims to understand and analyze the user-technology interaction and technology-engagement success measures to assess online customer engagement with chatbots and the impact on repurchase intention, within e-commerce websites. The sample data consists of 227 online consumer responses collected through an electronic survey. Only 165 respondents, which have used a chatbot to assist the online purchase process, are included in the effective sample. This research contributes to the digital marketing literature by complementing existing research exploring human-technology interactions, assessing how consumers interact with chatbot technology and how it affects customer engagement and behavioral outcomes within e-retail contexts. The study findings provide several challenges for managers. Finally, it discusses emerging trends in the digital marketing field, offering insights for future research avenues.

Abstract

Abstract
Energy hub systems improve energy efficiency and reduce emissions due to the coordinated operation of different infrastructures. Given that these systems meet the needs of customers for different energies, their optimal design and operation is one of the main challenges in the field of energy supply. Hence, this paper presents a two-stage stochastic model for the integrated design and operation of an energy hub in the presence of electrical and thermal energy storage systems. As the electrical, heating, and cooling loads, besides the wind turbine's (WT's) output power, are associated with severe uncertainties, their impacts are addressed in the proposed model. Besides, demand response (DR) and integrated demand response (IDR) programs have been incorporated in the model. Furthermore, the real-coded genetic algorithm (RCGA), and binary-coded genetic algorithm (BCGA) are deployed to tackle the problem through continuous and discrete methods, respectively. The simulation results show that considering the uncertainties leads to the installation of larger capacities for assets and thus a 8.07% increase in investment cost. The results also indicate that the implementation of shiftable IDR program modifies the demand curve of electrical, cooling and heating loads, thereby reducing operating cost by 15.1%. Finally, the results substantiate that storage systems with discharge during peak hours not only increase system flexibility but also reduce operating cost.

Abstract
Maritime transportation has been, historically, a major factor in economic development and prosperity since it enables trade and contacts between nations. The amount of trade through maritime transport has increased drastically; for example, about 90% of the European Union’s external trade and one-third of its internal trade depend on maritime transport. Major ports, typically, incorporate multiple terminals serving containerships, railways, and other forms of hinterland transportation and require interterminal and intraterminal container transport. Many factors influence the productivity and efficiency of ports and hence their economic viability. Moreover, environmental concerns have been leading to stern regulation that requires ports to reduce, for example, greenhouse gas emissions. Therefore, port authorities need to balance economic and ecological objectives in order to ensure sustainable growth and to remain competitive. Once a containership moors at a container terminal, several quay cranes are assigned to the ship to load/unload the containers to/from the ship. Loading activities require the containers to have been previously made available at the quayside, while unloading ones require the containers to be removed from the quayside. The containers are transported between the quayside and the storage yard by a set of vehicles. This chapter addresses the intraterminal container transport scheduling problem by simultaneously scheduling the loading/unloading activities of quay cranes and the transport (between the quayside and the storage yard) activities of vehicles. In addition, the problem includes vehicles with adjustable travelling speed, a characteristic never considered in this context. For this problem, we propose bi-objective mixed-integer linear programming (MILP) models aiming at minimizing the makespan and the total energy consumption simultaneously. Computational experiments are conducted on benchmark instances that we also propose. The computational results show the effectiveness of the MILP models as well as the impact of considering vehicles with adjustable speed, which can reduce the makespan by up to 16.2% and the total energy consumption by up to 2.5%. Finally, we also show that handling unloading and loading activities simultaneously rather than sequentially (the usual practice rule) can improve the makespan by up to 34.5% and the total energy consumption by up to 18.3%. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Abstract
Capturing complex spatio-temporal features of thousands of correlated taxi-demand time-series in the city makes the traffic flow prediction problem a challenging task. Hence, several Deep Neural Network (DNN) models have been developed to mimic the latent spatio-temporal behaviour of taxi-demand time-series in a city to improve the prediction results. Despite, good performance of recent DNN based traffic prediction techniques, such models can only identify either adjacent or connected regions with direct or transitive connection; hence they fail to capture spatio-temporal correlation among regions that exhibit implicit or latent connection. Additionally, the dependency of the recent DNN models on recursive components facilitates error propagation during feature aggregation without any counter strategy for it. In view of these existing glitches, we introduce a novel DNN model, graph Multi-Head Convolution for Spatio-Temporal Aggregation (gMHC-STA) which supports capturing spatio-temporal correlation among regions with explicit and implicit connection both. Moreover, gMHC-STA aggregates both spatial and temporal characteristics using multi-head attention; thus overriding recursive RNN or its variant approach to prevent noise propagation. The experimental results of gMHC-STA on two real-world city taxi-demand datasets report minimum of 6.5–10% improvement over the best state-of-the-art on standard benchmark metric in varying experimental conditions. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Abstract
To automatically trap, manipulate and probe physical properties of micron-sized particles is a step of paramount importance for the development of intelligent and integrated optomicrofluidic devices. In this work, we aim at implementing an automatic classifier of micro-particles immersed in a fluid based on the concept of optical tweezers. We describe the automation steps of an experimental setup together with the implemented classification models using the forward scattered signal. The results show satisfactory accuracy around 80% for the identification of the type and size of particles using signals of 250 milliseconds of duration, which paves the path for future improvements towards real-time analysis of the trapped specimens.