联系我们

论文查询结果	返回搜索

论文编号：	15036
作者编号：	1120201078
上传时间：	2024/12/9 17:59:33
中文题目：	电商平台用户在线参与行为影响因素研究
英文题目：	Research on the antecedents of online user engagement in e-commerce platforms
指导老师：	李凯
中文关键字：	用户在线参与行为；评论内容丰富程度；情感；认知负荷；拟人化
英文关键字：	User Engagement； Review Content Richness； Emotion； Cognitive Load； Anthropomorphism
中文摘要：	作为用户在线活动的重要形式，用户在线参与行为的重要性被进行了广泛地讨论，尤其是在电子商务情景下。为了促进用户实施在线参与行为，现有文献从信息特征的角度讨论了用户在线参与行为的影响因素。然而，互联网信息的复杂性增加了学者们理解信息特征与用户在线参与行为关系的难度。为此，本研究以信息特征为切入点，通过对“信息特征—情感（认知）导向的信息处理—用户在线参与行为”这一递进逻辑进行拆解，建立了信息特征对用户在线参与行为的影响模型。通过分析来自于京东电商、TripAdvisor以及通过在线实验搜集的数据，本研究得到如下关键发现以及创新点： 1）本研究首先根据“信息特征—用户在线参与行为”这一逻辑讨论了评论内容丰富程度对用户投票行为的影响。使用自然语言处理技术，本研究提出了评论信息量的衡量方法，即评论内容丰富程度。该方法的提出对于丰富评论信息量相关文献有着十分重要的作用。此外，本研究检验了评论内容丰富程度与用户投票行为之间的倒U型关系，丰富了用户投票行为影响因素相关文献。最后，本研究还讨论了在正面、负面评分不一致情境下，评论内容丰富程度对用户投票行为的异质化影响，从而加深了现有研究对评论信息量与用户投票行为关系的理解。 2）根据“信息特征—情感导向的信息处理—用户在线参与行为”这一逻辑链条，本研究验证了用户感知到的信息量（即情感导向的信息处理结果）对于正面、负面情感信息和用户转发行为关系的中介作用。尽管相关研究讨论了情感信息对于用户转发行为的影响，但是鲜有研究着重分析正面、负面情感信息的异质化影响。通过捕捉评论内容中蕴含的情感信息数量，本研究分析了正面、负面情感信息数量对于用户转发行为的异质化影响，从而补足了现有研究的缺陷。除此之外，本研究还分析了正面、负面情感信息离散程度（个体和群体之间情绪的差异）和情感信息不一致程度（评论标题情感与内容情感之间的不一致）对于上述关系的调节作用，这在以前的研究中未被进行讨论。最后，本研究立足于用户感知到的信息量，揭示了正面、负面情感信息对用户转发行为的作用机制，从而加深了现有研究对于情感信息与用户转发行为关系的理解。 3）本研究依照“信息特征—认知导向的信息处理—用户在线参与行为”讨论了由评论特征诱发的不同类型的认知负荷对于用户在线参与行为的影响以及作用机制。首先，本研究是首批讨论不同类型认知负荷对用户在线参与行为影响的研究之一，从用户在线参与行为影响因素的角度丰富了用户在线参与行为文献。其次，本研究还考察了不同类型的认知负荷与用户在线参与行为之间的关系如何随着跨信息的认知负荷的变化而改变。这一研究超越了单一信息且立足于信息集的层次，为认知负荷和用户在线参与行为关系的文献做出了贡献。最后，本研究立足用户感知到的疲劳（即认知导向的信息处理结果），揭示了不同类型的认知负荷对用户参与行为的作用机制，深化了现有研究对于认知负荷与用户参与行为关系的理解。 4）本研究整合上述逻辑链条，讨论了在人工智能客服情境下，用户满意度（即情感导向的信息处理结果）和心理距离（即认知导向的信息处理结果）对于不同类型拟人化与用户信息分享行为关系的中介作用。首先，本研究通过讨论不同类型的拟人化对于用户信息分享行为的非线性影响，矫正了现有研究对两者关系过度简化的理解，为信息分享行为的影响因素相关文献做出了贡献。除此之外，本研究考虑到了用户满意度以及用户心理距离对于上述关系的中介作用，从而扩充了拟人化与用户在线信息分享行为关系的相关文献。本研究结论为拟人化界面设计者提供了具体的设计思路，并且为其聚焦设计重点提供了启示。
英文摘要：	As an important type of users’ online activities, the significance of online user engagement behaviors has been widely discussed, especially in the context of e-commerce. To promote user online engagement behaviors, the existing literature has explored the determinants of such behaviors from the perspective of information characteristics. However, the complexity of online information increases the difficulty for scholars to understand the relationship between information characteristics and user engagement. Therefore, this study takes information characteristics as the starting point and, by deconstructing the progressive logic of "information characteristics—emotion (or cognition)-oriented information processing—user engagement", establishes a model of how information characteristics influence user engagement. By analyzing the data collected from JD.com, TripAdvisor, and online experiments, this study presents the following key findings and contributions: Based on the logic of "information characteristics—user engagement", this study discusses the impact of review content richness on user voting behavior. Using natural language processing techniques, this study proposes a method to measure the amount of review information, namely review content richness. The proposal of this method plays a significant role in enriching the literature related to review information quantity. Additionally, this study examines the inverted U-shaped relationship between review content richness and user voting behavior, contributing to the literature on factors influencing voting behavior. Finally, this study explores the heterogeneous effects of review content richness on user voting behavior in the context of positive and negative rating inconsistency, thereby deepening the understanding of the relationship between review information quantity and voting behavior. Based on the logical chain of "information characteristics—emotion-oriented information processing—user engagement", this study validates the mediating role of users’ perceived information quantity (i.e., the result of emotion-oriented information processing) in the relationship between positive and negative emotional information and users’ sharing behavior. Although related research has discussed the effect of emotional information on users’ sharing behavior, few studies have focused on the heterogeneous effects of positive and negative emotional information. By capturing the amount of emotional information embedded in review content, this study analyzes the heterogeneous impacts of positive and negative emotional information on users’ sharing behavior, thus addressing gaps in the existing literature. Moreover, this study examines the moderating effects of emotional information dispersion (variance in emotions among individuals or groups) and emotional inconsistency (incongruence of emotion between review titles and content) on the aforementioned relationships, which have not been discussed in previous research. Finally, this study, based on users’ perceived information quantity, reveals the mechanism of how positive and negative emotional information influence users’ sharing behavior, thereby deepening the understanding of the relationship between emotional information and sharing behavior. Based on the logic of "information characteristics—cognition-oriented information processing—user engagement", this study discusses the impact and mechanism of different types of cognitive load induced by review characteristics on user engagement. First, this study is among the first to discuss the effects of different types of cognitive load on users’ online engagement behaviors, enriching the literature on factors influencing online engagement behaviors. Second, this study also examines how the relationship between different types of cognitive load and users’ online engagement behaviors changes with variations in cross-information cognitive load. This research goes beyond individual information and operates at the level of information sets, contributing to the literature on the relationship between cognitive load and users’ online engagement behaviors. Finally, based on users’ perceived fatigue (i.e., the result of cognition-oriented information processing), this study reveals the mechanism of how different types of cognitive load influence users’ engagement behaviors, thereby deepening the understanding of the relationship between cognitive load and user engagement. This study integrates the above logical chains to discuss the mediating roles of user satisfaction (i.e., the result of emotion-oriented information processing) and psychological distance (i.e., the result of cognition-oriented information processing) in the relationship between different types of anthropomorphism and users’ self-disclosure behaviors in AI chatbot contexts. First, by discussing the nonlinear effects of different types of anthropomorphism on self-disclosure behaviors, this study corrects the overly simplistic understanding of the relationship in existing research, contributing to the literature on factors influencing self-disclosure behaviors. Moreover, this study considers the mediating roles of user satisfaction and psychological distance in the aforementioned relationships, thereby expanding the literature on the relationship between anthropomorphism and self-disclosure behaviors. The conclusions of this study provide concrete design strategies for designers of anthropomorphic interfaces and offer insights for prioritizing areas of focus in design.
查看全文：	预览  下载(下载需要进行登录)