Using firm-level data on agricultural enterprises from 2010 to 2022, this paper employs the staggered rollout of Chinese “Comprehensive Demonstration of E-commerce Entering to Rural Areas” Program as a quasi-natural experiment to identify the impact of rural e-commerce (REC) policy on agricultural corporate performance. The results show that REC policy significantly enhance agricultural enterprise performance. Mechanism analyses indicate that the effect operates through digital transformation and market advantages. Heterogeneity tests reveal that this impact is more significant for private and small agricultural enterprises, for those in eastern regions or where local governments pay more attention to the digital economy. This research contributes to the comprehension of the micro-level effects of rural e-commerce and offers beneficial suggestions for the high-quality development of agricultural enterprises.
Conversational AI systems are increasingly common on e-commerce platforms, where they provide instant product consultation and personalized recommendations. Graph-based retrieval-augmented generation (RAG) frameworks have further advanced these systems by enabling more natural and contextually aware responses. In this paper, we present LibraAI: a lightweight conversational AI designed to help users find suitable books on Tiki, a Vietnamese e-commerce platform with a highly structured and specialized book catalog. LibraAI is built on an enhanced graph-based RAG framework and introduces a novel relation-oriented retrieval method. Our experiments show that LibraAI reduces input-context token usage by 67
In global B2B e-commerce, geopolitical disruptions impose a rigid financing dilemma on suppliers: either bear escalating risk premiums on purchase-order loans or lock in unconditional factoring discounts, irrespective of whether a shock materializes. We develop a two-stage game-theoretic model to examine how open finance, through real-time APIs embedded in e-commerce platforms, enables flexible factoring that activates exclusively upon verified shock realization. We demonstrate that traditional purchase-order financing and non-flexible factoring exhibit a risk-dependent trade-off, each dominating only under low or high disruption probabilities. Flexible factoring eliminates this trade-off entirely. By applying discounts solely to the residual maturity of receivables and removing the bank’s geopolitical risk premium, it strictly dominates both static instruments, raising order quantities and profits for all supply chain partners at every risk level. Extensions to retailer-side and bilateral risks further confirm that this advantage persists across risk contexts, establishing real-time optionality as a robust mechanism for building digitally integrated resilience in e-commerce supply chains.
Helpful online reviews play a critical role in influencing consumer decision-making and driving sales on e-commerce platforms. While scholarly interest in the influence of emotional cues on review helpfulness has increased, extant research predominantly conceptualizes emotional content through broad negative or positive sentiment categories. This approach overlooks the role of discrete emotions, such as anger, fear, disgust, and sadness. Although these emotions share a negative valence, they differ in their psychological antecedents and in how they shape consumer interpretation. Treating them as a uniform category limits both theoretical nuance and practical insight into the mechanisms underlying review helpfulness. Drawing on emotional appraisal theory, we explore the direct and interactive impacts of negative discrete emotions on review helpfulness and the moderating impact of review image colorfulness. Using Amazon reviews as a data source, we find that anger enhances review helpfulness, while sadness and disgust diminish it. Additionally, review image colorfulness amplifies the effects of anger and disgust on review helpfulness. Theoretically, this study advances the literature by distinguishing the effects of negative discrete emotions and highlighting the role of image colorfulness in moderating the emotional impact. It also offers novel insights for platform designers and marketers by demonstrating how specific emotional expressions and visual cues jointly shape consumers’ perceptions of review helpfulness.
Information asymmetry is a common phenomenon that exists in two-sided markets. Participants in two-sided markets tend to hide their true information out of self-interest. To explore the effect of platform’s quality effort cost misreporting (QECM) and users’ fairness concerns misreporting (FCM) on optimal platform decisions in two-sided markets, this paper constructs four Hotelling’s bilateral platform pricing models under two dimensions of information asymmetry. By analyzing the equilibrium results, we conclude some important findings. Firstly, users’ FCM causes non-neutral of platform pricing, i.e., the platform skews pricing to side with misreporting behavior. Secondly, both platforms’ QECM and users’ FCM exacerbate the potential moral hazard problem, and the platform’s quality effort level reaches a minimum state when both platforms and users misreport their own private information. Finally, we find that misreporting of information by both platforms and users exploits information asymmetry as a self-interest and detriment to others strategic behavior. Intrinsically, platforms balance the two-sided system by establishing a cross-subsidy transfer mechanism that absorbs the impact of behavioral misreporting. Through numerical simulations, we verify that FCM exacerbates non-neutral pricing structures. Furthermore, both misreporting behaviors lower platform quality effort, with the lowest quality effort observed under simultaneous misreporting.
