With the growing attention to information security (infosec) in safeguarding data assets, more and more firms voluntarily disclose their infosec practices to cater to the market. Such disclosures not only affect the trust of investors but also have significant impacts on hackers’ behaviors. Previous studies empirically discussed how infosec disclosure affects firm value, but didn’t consider the authenticity of such disclosure. This study develops a game-theoretic model involving one firm and a strategic hacker, aiming to analyze how overstatement in infosec disclosure affects both firms and hackers. We find that when the firm discloses infosec practice, compared with the non-disclosure situation, the threat faced by the firm will decrease but the breach probability will increase, while the hacker’s effort and expected benefit will increase. When the firm overstates infosec practices, compared with honest disclosure, the threat and the breach probability will decrease, while the hacker’s effort and expected benefit will decrease. Moreover, we find that when the firm overstates the infosec practice, it increases other firms’ expected costs, resulting in a negative spillover effect in the industry.
The detection of material weaknesses in internal control (MWIC) of companies is essential since it provides early warning for various financial risks. Currently machine learning methods have been extensively adopted in this field. However, previous studies have not fully captured the relationships among different modalities, and neglected the consistency and complementary information. To address these challenges, we propose a Consistency and Complementary information-aware Multi-modal Deep Learning method (namely, CCMDL), which effectively captures the consistency and complementary relationships among different modalities. CCMDL comprises three main components, namely, the multi-modal feature extraction, the multi-modal feature enhancement and multi-modal feature fusion. Initially, multi-modal feature extraction module extracts valuable features and transforms them into deep representations. Subsequently, multi-modal feature enhancement module captures both the consistency and complementary information between these modalities, leveraging the joint effect of multi-modalities for MWIC detection. Finally, multi-modal feature fusion module introduces a novel pooling mechanism to detect MWIC. Through the experimental results on the real-world dataset, the proposed CCMDL outperforms the benchmark method in detecting MWIC. Our method enhances the effectiveness of detecting MWIC and provides robust decision-making support for regulators and investors.
After the sudden disaster occurred, the damage to communication infrastructure and road capacity can lead to an explosive increase in demand for emergency supplies, and the demands for different types of emergency supplies can vary depending on the period. Due to the poor communication between the affected area and the external environment, the demand generation pattern will be unpredictable. To address this challenge, we propose a novel anticipatory routing, acceptance, and postponement policy for the Multi-period Dynamic Vehicle Routing Problem with Stochastic Requests (MDVRPSR). We first provide a detailed description of the unique features of MDVRPSR, followed by the formulation of a mathematical model using Markov Decision Process (MDP). Our model is designed to respond to stochastic requests within a multi-period dynamic framework, offering a comprehensive perspective on the decision state, reward setting, transition process, and objective function. Subsequently, we employ Deep Reinforcement Learning (DRL) to optimize the routing policy. Experiments demonstrate that the DRL-based policy improves the effectiveness of emergency supplies distribution in dynamic and uncertain scenarios.
With the rapid growth of short video platforms, covert advertising has emerged as a key strategy for driving consumer behavior in e-commerce. However, existing research has primarily focused on explicit advertising and its effects on consumer attitudes, with limited exploration of how covert advertising in short video e-commerce influences impulse buying through psychological mechanisms like perceived entertainment and trust. To address this gap, this study investigates the impact of covert advertising features—creativity, content-congruity, and personalization—on consumers’ impulse buying intentions, based on the Stimulus-Organism-Response theory. Data were collected via a survey of active short video platform users and analyzed using structural equation modeling. This study extends the applicability of the Stimulus-Organism-Response framework in digital marketing contexts and offers practical insights for optimizing short video advertising design to enhance consumer engagement and conversion rates.
To investigate the pricing and power battery warranty decisions of new energy vehicle enterprises when the consumers have different perceptions of the battery warranty service, we consider a supply chain system in which a battery warranty service is provided by a new energy vehicle manufacturer. We first derive the optimal decisions of the new energy vehicle enterprise using the Nash equilibrium. Then, we conduct the sensitivity analysis on some critical parameters and compare the optimal profits of the new energy vehicle enterprise in different scenarios. The results show that the consumers’ sensitivity coefficient to the battery warranty service and the scaling parameter for battery failure distribution are related to the decision making of new energy vehicle enterprises. Based on the above study, we provide theoretical support for new energy vehicle enterprises to make more appropriate and profitable decisions.
This study explores the differences between AI streamers and human streamers in live-streaming e-commerce and examines how these differences influence consumer purchase intention. The study employs the Stimulus-Organism-Response model and adopts a case study approach, conducting semi-structured interviews with 28 young adults. The collected data are analyzed through open coding, axial coding, and selective coding. The findings reveal that human streamers outperform AI streamers in sensory, functional, and affective dimensions, such as sound quality, knowledge reserve, and emotional disclosure. These differences impact consumer perceived trust, which in turn affects purchase intention. Furthermore, consumer purpose (entertainment vs. shopping) moderates the role of perceived trust in shaping purchase decisions. This study enriches and expands existing research on purchase intention in live-streaming e-commerce by comparing the influence mechanisms of AI and human streamers, providing new evidence for the comparative study of their impact on consumer behavior.
Based on the relevant data of listed companies in the Chinese robotics industrial chain, this paper constructs an evaluation index system for the resilience of the Chinese robotics industrial chain from four dimensions, i.e., enterprise and enterprise cluster resilience, supply chain network resilience, industrial infrastructure resilience, and industrial dynamic development resilience. On this basis, this paper employs fuzzy comprehensive evaluation method to measure the importance of each first - level indicator to the Chinese robotics industrial chain resilience, and uses the entropy - weight VIKOR method to conduct a comparative analysis and ranking of the resilience of the robotics industrial chain in various provinces and cities. The results reveal that supply chain network resilience has the greatest impact on the robotics industrial chain resilience. Using the standard deviation method, provinces are classified into four tiers: Shanghai, Guangdong, Zhejiang, Beijing, and Jiangsu are at a high – resilience level, Shandong, Hubei, Fujian, Anhui, and Tianjin are at a medium – resilience level. Finally, this paper analyzes strengths and weaknesses of robotics industrial chain resilience in various provinces and cities.
With more and more fresh food e-commerce platforms to choose from, consumers are easy to switch platform, and it is extremely important to study the consumers’ intention to switch fresh food e-commerce platform. Based on the PPM theory, this paper integrates push factors (price, quality), pull factors (alternative attractiveness, word of mouth) and mooring factors (switching cost, subjective norms, variety seeking), and uses fsQCA method to explore the antecedent configuration path of consumers’ intention to switch fresh food e-commerce platform. The results show that the single antecedent condition can not constitute the necessary condition of high switching intention; there are five equivalent paths for the generation of high switching intention, among which alternative attractiveness and variety seeking play a more universal role in generating high switching intention. This paper helps to reveal the complex causal mechanism of multiple factors jointly influencing consumers’ fresh food e-commerce platform switching intention, providing beneficial enlightenment for fresh food e-commerce enterprises to enhance the stickiness of consumers on the platform.
E-commerce emerges from the integration of internet technology and business models, shaped by technological advancements, market demand shifts, and policy changes, leading to distinct industry characteristics across eras. Upper Echelons Theory posits that entrepreneurs’ backgrounds and traits influence corporate performance in complex, multifaceted ways, necessitating non-linear research approaches. This paper employs the fs QCA method to explore the configuration characteristics of e-commerce entrepreneurs across different temporal contexts and their impact on performance. Analyzing 89 cross-sectional cases from 1996 to 2023, the study identifies diverse high-performance paths for entrepreneurs in different eras. Key findings include: First, high-performance paths exhibit significant diversity and era-specific characteristics. Second, risk propensity, business acumen, and technical background are core conditions for achieving high performance. Third, the policy environment and technological changes significantly shape entrepreneurial configurations. This study enriches individual-level research on entrepreneurs and offers insights for e-commerce companies in establishing and selecting leaders.
Studying the factors affecting users’ continuance intention on Meituan Youxuan platform is of vital importance to Meituan Youxuan platform operators, which helps to accurately grasp users’ behavior. Based on the Information System Continuous Use Model (ECM-ISC), this paper incorporates the two factors of trust and habit into the model to deeply explore their impact on users’ continuance intention. SPSS 26 and Amos 24 were used to conduct an empirical study on 220 valid questionnaires. The results show that: (1) expectation confirmation, perceived usefulness and trust have a significant positive impact on user satisfaction; (2) perceived usefulness, satisfaction, trust and habits all have a significant positive effect on users’ Continuance Intention. Based on this, according to the five aspects of expectation confirmation, perceived usefulness, satisfaction, trust and habits, specific suggestions are put forward for Meituan Youxuan platform operators to further enhance user loyalty.
The innovation capability of artificial intelligence (AI) enterprises serves as a significant driving force for the high-quality development of the AI industry. This paper takes 63 listed AI enterprises in China as research samples and constructs an analysis framework for the innovation performance of AI enterprises based on Technology-Organization-Environment (TOE) theory. Using fuzzy-set qualitative comparative analysis to examine the innovation capabilities of listed AI enterprises in China. The research reveals that no single factor is necessary for high innovation performance among AI enterprises. Instead, there are four configurational pathways to improve AI enterprises innovation performance, which can be summarized into three types: technology-organization-environment-driven type, technology-organization-driven type, and technology-environment-driven type. Based on objective data, The conclusion of the study reveals the complex mechanism of multiple concurrent factors on the formation of high innovation performance of AI enterprises, provides theoretical support for enterprises to promote innovation and achieve high-quality development, and provides practical reference for the government to optimize policies to support AI enterprises.
Trade credit, as a form of supply chain finance, can alleviate the financing pressure of small and medium-sized enterprises. However, it also presents certain drawbacks, such as the risk of retailer bankruptcy or delayed payments. The introduction of blockchain (BCT) and artificial intelligence (AI) technologies can ensure that retailers repay their debts promptly, and offer additional benefits, such as enhancing consumer trust in products, improving the accuracy of predictions, and reducing the costs associated with BCT usage. This paper constructs a supply chain consisting of a financially constrained retailer and a supplier to analyze and model the decision-making processes under four scenarios: with or without the use of BCT and AI. Through calculations, it is found that the integration of BCT and AI into the supply chain significantly influences the decision-making of both retailers and suppliers.
Bilibili’s scientific popularization sub-sectors under the knowledge category provides an excellent learning platform for a broad audience. However, some science popularization videos on the platform suffer from low popularity. Research on how to identify scientifically influential videos remains underexplored. Against this backdrop, this research defines a video influence calculation method based on entropy weight, tailored to Bilibili’s specific characteristics, and constructs a Multimodal CNN-Transformer model (MMCT) that integrates text, social, audio, image, and content modalities to predict the influence of scientific popularization videos on Bilibili. Experiments are conducted on video influence regression prediction using crawled Bilibili video data, and the model’s effectiveness is validated through ablation studies. Furthermore, benchmarks are set to verify the effectiveness of the multimodal feature fusion in video influence classification. The results indicate that MMCT achieves an SRCC of 0.7895, improving by 0.0566 compared to the baseline model, while MSE decreases by 0.4022 compared to the baseline model. This study contributes to Bilibili’s ability to identify highly influential popular science videos, thereby enhancing the platform’s commercial value and competitiveness while promoting the effective and high-quality dissemination of popular science knowledge on Bilibili.
Online team consultations (OTCs) enhance existing online healthcare services by providing more comprehensive care compared to consultations with a single physician. However, poor teamwork quality, such as a lack of coordination and uneven contributions, can result in patients discontinuing the use of online medical team (OMT) service. This not only disrupts the development of long-term patient relationships but also diminishes the overall benefits of collaborative care. Therefore, the aim of this study is to explore how the quality of teamwork, specifically coordination and balance of contributions, within OMTs influences patients’ willingness to continue using team-based consultation services, while also investigating the moderating role of teamwork experience through the Stimulus-Organism-Response (S-O-R) framework. Our research model was validated using public data collected from a leading online health platform in China. We find that coordination and balance of contributions positively and significantly impact patient satisfaction, which subsequently leads to continuous use. Teamwork experience strengthens the effect of the balance of contributions on satisfaction, but has an insignificant influence on the link between coordination and satisfaction. This study deepens the understanding of how patients perceive OMT teamwork quality and their decision-making process about continued OTC service use.
With the rapid development of artificial intelligence technology, conversational AI is increasingly widely used in daily life. However, technical limitations and scenario complexity make AI service failure inevitable. After service failure, how to repair it effectively becomes the key to maintaining user trust. This study takes conversational AI as the research object to explore the impact of different apology strategies and service types on user forgiveness. Through two experiments, the results show that general humility and intellectual humility positively affect user forgiveness; Compared with no apology, both can significantly improve user forgiveness; Service type moderates the effect of apology strategy. In tool-type service, an intellectual humble apology is more effective, while in chat-type service, a humble apology is more favored by users. The results of this study provide empirical evidence and practical guidance for developers of conversational AI to optimize user experience and improve user satisfaction. Developers need to pay attention to users’ perceptions and understandings of different apology strategies and constantly optimize their design and expression to align with users’ expectations and emotional needs.
This paper develops an equilibrium model to investigate pricing and technology upgrade decisions in the context of supply chain versus supply chain competition. The model considers a two-echelon supply chain where manufacturers and retailers make decisions on pricing, output, and artificial intelligence technology upgrades. Demand is influenced by both price and service level, and technology upgrades implicitly enhance customer utility by improving the service level. These interdependencies across the supply chain significantly impact overall profitability. Consequently, pricing and technology upgrade decisions must be optimized collectively rather than in isolation. Three key features of this model are highlighted: First, our model is built upon the context of supply chain versus supply chain competition, rather than the context of firm versus firm competition. Second, it integrates pricing and technology upgrade decisions to enhance the overall performance of a supply chain network. The interplay between pricing strategies, technological advancements, and customer utility is analyzed to improve supply chain efficiency and competitiveness. Third, the model assumes that heterogeneous customers are sensitive to both price and service level, with technology upgrades playing a pivotal role in shaping customer preferences. We utilize variational inequalities to characterize the equilibrium conditions of the supply chain, enabling robust analysis and decision-making. This study provides a novel framework for understanding pricing and technology upgrade strategies in a competitive supply chain network, offering actionable insights for decision-makers seeking to optimize supply chain performance and enhance customer satisfaction in competitive markets.
This study establishes a comprehensive classification system for user search assistance needs through extensive research on existing studies and in-depth interviews with search users. Subsequently, based on the proposed auxiliary demand classification, three machine learning algorithms and a deep learning algorithm are applied to construct a classification prediction model, which implements a classification prediction method from user search behavior characteristics to user search auxiliary demand. The experimental results show that the random forest model exhibits the best performance in the prediction problem, and its average prediction accuracy and AUC index for the six categories of search assistance auxiliary needs are close to 0.9.
This study addresses the challenge of exploring the dynamics of cross-organizational data sharing processes. The approach begins by using BPMN to model cross-organizational business processes. It then employs CPN Tools for formal modeling and simulation, thereby establishing a formal analysis framework for capturing the complexities inherent in dynamic collaborative data-sharing processes. Focusing on the railway industry, the study uses the case of handling obstructions encountered by the pantograph during high-speed train operations. This scenario exemplifies railway multi-Disciplinary collaborations, involving key stakeholders such as train operators, dispatch centers, and power supply departments. The process is characterized by stringent timeliness requirement sand complex cross-departmental collaboration. The study provides valuable insights for enhancing cross-organizational coordination efficiency and improving data sharing among multiple stakeholders.
Native advertising and display advertising are often combined as each serves a distinct business purpose. However, their influential relationship has not been thoroughly examined, which may blind platforms to making sound advertising portfolio decisions. This study explores the causal effect of display advertising on native advertising performance in mobile contexts, and the moderating roles of sellers’ (i.e., restaurants’) average rating and chain status. A field quasi-experiment on a large mobile e-commerce platform shows that following the introduction of display ads, the clicks, bids, and costs of native ads decreased significantly by 6.6
[Objective] This paper is an empirical study aimed at gaining insights into identifying the instinctive information demands of individuals involved in or impacted by ADHD (attention-deficit/hyperactivity disorder), including those caring for individuals with ADHD, to engage in online health community activities. We accomplish this work by collecting, cleaning, and analyzing user-generated posts in ADHD OHCs (online health communities). [Design/methodology/approach] This study uses Self-Determination Theory (SDT) to explore users’ instinctive motivations. This paper introduces the classical Latent Dirichlet Allocation (LDA) as a text-mining method to analyze and categorize topics. In the process, this study conducts a joint consistency and perplexity test on the clustering results to detect the optimal number of topics and improve the accuracy of topic identification. In addition, this study mapped the obtained topic clustering results to four main scenarios, including the online health community homepage, topic detail page, personal information page, and posting page, to explore users’ health information demands in different scenarios. [Results] The results suggest that text mining methods can help identify user demands related to health and platforms and enable a more comprehensive assessment of the information demands of users with ADHD. [Originality/value] This study uses text mining methods to summarize the vast amount of text content in OHCs, further enriching our understanding of the types of user information demands. Applying the SDT theory to the motivation assessment of ADHD patients enhances the interpretability of the text clustering results. It helps discover online users’ motivations and preferences when engaging in community activities.
