Electronic Commerce Research最新文献

As image-based communication proliferates, the business value of image sentiment analysis is rapidly growing, particularly in fields like advertising where consumers receive emotional cues through visual stimuli. However, most existing research on image sentiment analysis has focused more on developing sentiment classification models rather than exploring specific factors contributing to image sentiment. Therefore, this study proposes a methodology for extracting color palettes to represent image sentiments, emphasizing the significance of color palettes as highlighted in various studies. Previous approaches to color palette extraction have included heuristic methods, survey-based selection, or utilizing clustering algorithms like K-means clustering based on color frequencies in images. In this study, we calculate the influence scores of colors for classifying image sentiments and propose deriving representative sentiment-color palettes based on these scores. Initially, we train a multi-label classification model to predict the sentiment labels of images and then create datasets for distorted images where pixels corresponding to specific colors are removed. By comparing the model outputs obtained from these distorted images with the original dataset, we obtain quantitative influence scores of colors for classifying sentiment labels. Furthermore, we extract sentiment-color palettes consisting of four important colors for 30 different sentiments. Experimental results demonstrate higher evaluation scores compared to previous studies.

To cope with the challenge caused by the lack of physical evaluation, many e-commerce platforms introduce a showroom channel, in which consumers can feel and touch products before purchasing. Based on this background, we consider there are two strategies of an e-commerce platform to introduce a showroom channel, i.e., self-build strategy and cooperation strategy. Then we build theoretical models by characterising cost inefficiency of self-build strategy and information asymmetry of cooperation strategy in the platform supply chain. Next, we derive the optimal decisions under different strategies, and investigate the advantage of different strategies. Last, we explore the strategic introduction of the showroom channel in the platform supply chain. We find the following results. (1) Under cooperation strategy, the platform can strategically design two different cooperation contracts to the retailer, i.e., full incentive contract and partial incentive contract. Moreover, the retailer may be hurt by a higher demand due to information asymmetry under cooperation strategy. (2) Self-build strategy generates information advantage and cooperation strategy brings channel advantage, which jointly affect the optimal showroom channel strategy of the platform. To be more specific, when high-type demand of the retailer is relatively high and reservation profit of the retailer is relatively low, channel advantage dominates information advantage, then the platform will choose cooperation strategy. (3) Under some conditions, the optimal strategy of the platform is not in line with the optimal ones of the whole supply chain. Then we design a side payment contract to achieve Pareto improvement.

Rural tourism plays an important role in advancing the rural economy but faces unexpected risks and challenges such as the deterioration of operational performance. Information and communications technologies (ICT) – such as digital payment platforms, online booking systems, social media marketing, and virtual reality tours – become pervasive in the tourism industry during the current digital era. However, little empirical evidence has been presented about how such technologies impact rural tourism and if rural tourism resilience gets enhanced with the support of ICT in China. This study addresses this research gap by analyzing a sample of 20,716 rural tourism operators from the Third National Agricultural Census of 2016 in Beijing, which is a major urban center in China with a substantial market for rural tourism products and services. With the aid of a treatment effect model (TEM), the research results show that ICT can be an effective way to promote the rural tourism industry through mitigating business difficulties and risks and therefore to enhance the resilience of rural tourism. Our study shows that ICT plays an important and positive role in promoting tourism sales externally and enhancing management internally. The heterogeneity of the operator’s organization type and human capital also makes a difference. The sample data robustly show that ICT can effectively alleviate “scale discrimination” and support “low education level bias”, which is conducive to small-scaled business households and has also increased the resilience of those operators who are at a disadvantaged position in the industry.

Collaboration and value co-creation are important drivers of the continuous growth of e-commerce, which is expected to reach US $6.4 trillion in 2024 despite current global crises. Only a few transaction platforms currently dominate e-commerce (eg., Amazon, Walmart), but other participants are likely to join these platforms’ ecosystems. Third-party developers can provide extensions to these ecosystems to enhance the platforms’ functionality, but third-party developers’ role in e-commerce ecosystems’ success and generativity remains underexamined in academia. The present study scrutinizes the efficacy of boundary resources in attracting and managing third-party developers in e-commerce ecosystems. This investigation is predicated upon qualitative data gathered through interviews with 14 domain experts. The insights derived from these interviews have culminated in the formulation of seven design principles. These design principles are envisaged to serve as a guiding framework for owners of innovation and transaction platforms within the e-commerce sphere, facilitating the strategic deployment of boundary resources. It is anticipated that collaboration, value creation, and the overall generative capacity as well as the success of e-commerce ecosystems shall be considerably enhanced.

In the era of invasive social media and advanced artificial intelligence, sentiment analysis has become a vital tool for e-commerce and businesses to grasp user needs and monitor brand perception. This is particularly relevant in the film industry, where understanding the determinants of a movie’s pre-release performance is crucial for producers and investors. Traditional methods often rely on complex algorithms that lack transparency in elucidating the relationship between key risk factors and movie outcomes. This study addresses this gap by employing an explainable analytics framework to investigate the impact of various social media post characteristics on movie performance before its release. Initially, an exploratory data analysis was undertaken to identify significant risk factors associated with movie failures. Subsequently, the study segmented the analysis into three risk categories—low, moderate, and high risk—and applied conventional machine learning models to forecast the likelihood of failure within each category. The culmination of this research involved the application of a SHapley Additive exPlanation (SHAP) model, which provided insightful interpretations of how different risk factors contribute to the potential success or failure of movies. By integrating SHAP for interpretability, this research offers novel insights into the predictive dynamics of movie performance, paving the way for informed decision-making in the film industry.

This paper explores the role of artificial intelligence (AI)-powered voice assistants (VAs) in the context of online shopping, with a specific focus on Indian consumers. Through a quantitative research approach, data were collected with an online survey of 150 Indian participants based on constructs and measurement tools, which are meticulously defined, ensuring the reliability and validity of the results. The study (1) explores previous research to understand the use of AI-powered VAs in online shopping and their varied antecedents/dimensions; (2) analyses the influence of consumer trust on intention to use, satisfaction, and emotional attachment among individuals who use AI-powered VAs in online shopping; (3) develops a framework that reflects the antecedents of AI-powered VAs in online shopping and the outcome as brand loyalty, while taking consumer trust as the mediator for users’ intention to use AI-powered VAs, customer satisfaction, and emotional attachment in their journey towards brand loyalty; (4) studies how Alexa, an AI-powered VA, influences the consumer journey when shopping online in India and ultimately affects the brand loyalty of Indian consumers.

An increasing number of rewarded crowdfunding platforms recommend that fundraisers post multimodal data to improve data diversity and attract investors’ attention. Fundraisers’ speech contains both textual modality (text) and acoustic modality (voice) data. This study aims to explore how linguistic style of the speech text and acoustic features of the speech voice influence the crowdfunding campaign performance from the perspective of emotional contagion. An econometric model to investigate the effects of emotional cues from fundraisers’ speech on rewarded crowdfunding performance is constructed, and an empirical analysis with 21,996 projects data in Kickstarter is conducted. The findings demonstrate that both acoustic emotional cues (voice pitch, intensity and speech rate) and textual emotional cues (intimate and perceptual language) have significant effect on the financing performance. Among them, voice pitch, speech rate, intimate language and perceptual language have significant positive effect on the financing performance, while voice intensity is negatively related to financing performance. The number of comments positively moderates the relationship between emotional cues in speech and project financing performance. Research also finds that the acoustic and textual emotional features in speech have different effects on the success rate of project financing under different types of projects. This investigation provides both theoretical implications for the literature of crowdfunding and practical implications for fundraisers.

Incorporating fluency theory into a user experience design framework, this study utilises design considerations, drawing on complementary angles of the two theoretical domains, as a foundation with the ultimate goal of creating beneficial mobile payment experiences. An exploratory approach is deployed through semi-structured interviews to provide insights into experience design considerations utilising sensory elements and risk perception, a combination which has thus far received little attention. Research participants consist of senior managers that work for companies that create, facilitate, or accept mobile payment apps or processes. A conceptual framework is proposed with design as a starting point, including aesthetics and the need for a simplified experience, along with sensory elements that replicate familiar visual, audio, and haptic stimuli. These lead to a more usable experience that is perceived as easy to use through a frictionless experience. Usefulness is increased as exposure increases, and new app or process features can be added once prior features become familiar through repeated use. Key trade-offs include a simplified experience versus feature-rich experience, and frictionless experiences versus security risks, with key practical suggestions on how these can be approached.