电商情境下大模型驱动的对话式购物助手对持续使用意愿的影响研究——基于双重信任视角的实证分析

A Study on the Impact of Large Model-Driven Conversational Shopping Assistants on Consumers’ Continuous Usage Intention in E-Commerce Contexts—An Empirical Study Based on the Perspective of Dual Trust

摘要: 生成式人工智能(AIGC)与大语言模型的爆发正在重塑电子商务的交互范式。相较于传统智能客服，大模型驱动的对话式购物助手具备更强的拟人化交互与自然语言处理能力。本研究基于人机交互理论与双重信任框架，构建了探讨大模型技术特征(拟人化特征、感知智能性)对消费者持续使用意愿影响机制的理论模型。通过对356份有效问卷数据进行结构方程模型(SEM)分析，研究发现：大模型的拟人化特征能够同时显著增强消费者的认知信任与情感信任，而感知智能性则是构建消费者认知信任的核心来源；认知信任与情感信任在技术特征与持续使用意愿之间发挥了显著的双重中介作用，且在人机对话情境下，情感信任对持续使用意愿的驱动效应(0.415)强于认知信任(0.384)。本研究不仅拓展了对话式电商情境下大语言模型的应用边界，深化了消费者行为理论，也为平台企业优化AI助手的设计与留存策略提供了切实的管理启示。

Abstract: The outbreak of generative artificial intelligence (AIGC) and large language models is reshaping the interaction paradigm of e-commerce. Compared with traditional intelligent customer service, dialogue-based shopping assistants driven by large models have stronger anthropomorphism interactions and natural language processing capabilities. This study, based on human-computer interaction theory and a dual trust framework, constructs a theoretical model to explore the impact mechanism of large model technology features (anthropomorphism characteristics, perceived intelligence) on consumers’ continuous usage intentions. Through structural equation modeling (SEM) analysis of 356 valid questionnaire responses, the study finds that the anthropomorphism characteristics of large models can simultaneously significantly enhance consumers’ cognitive trust and affective trust, whereas perceived intelligence is the core source for building consumers’ cognitive trust. Cognitive trust and affective trust play a significant dual mediation role between technology features and continuous usage intention, and in human-computer dialogue contexts, the driving effect of affective trust (0.415) on continuous usage intention is stronger than that of cognitive trust (0.384). This study not only enriches the consumer behavior theory of large language models in e-commerce scenarios but also provides practical management insights for platform companies to optimize AI assistant design and retention strategies.