联系我们

论文查询结果	返回搜索

论文编号：	15343
作者编号：	2120223820
上传时间：	2025/6/6 10:23:33
中文题目：	AIGC服务质量对学术用户持续使用意愿的影响及优化策略研究
英文题目：	Research on the Influence of AIGC Service Quality on the Continuous Usage Intention of Academic Users and Optimization Strategies
指导老师：	樊振佳
中文关键字：	关键词：人工智能生成内容；期望确认模型；服务质量；持续使用意愿；学术用户
英文关键字：	Keywords:Artificial Intelligence-Generated Content; Expectation Confirmation Model; service quality; continuance usage intention; academic users
中文摘要：	人工智能生成内容（AIGC）技术的快速发展正在深刻改变学术科研的范式。随着深度学习、自然语言处理和大数据技术的突破，AIGC已从简单的文本生成工具演变为能够辅助文献综述、实验设计、数据分析甚至学术写作的智能伙伴。然而，尽管AIGC在学术场景中的应用日益广泛，其服务质量如何影响学术用户的持续使用意愿仍缺乏系统性研究。AIGC开发者在优化产品时也因此缺乏相应的理论支持，学术用户在使用过程中也面临信任度不足、黏性偏低等问题。因此，探索AIGC服务质量与学术用户持续使用意愿之间的作用机制，既是理论深化的必然要求，也是实践优化的迫切需求。 本文以期望确认模型为核心理论框架，结合服务质量理论，构建了包含有形性、信息质量、响应性、同理心、伦理安全五个维度的AIGC服务质量评价体系。研究通过三个阶段展开：首先，基于文献综述提炼初始模型，涵盖10个服务质量维度与30项具体指标；其次，采用半结构化访谈法对10名AIGC领域专家及长期使用AIGC的学术用户进行深度调研，结合质性分析优化模型，形成五维度（有形性、信息质量、响应性、同理心、伦理安全）共19项测量指标的修正框架；然后，通过问卷调查收集265份有效数据，利用结构方程模型验证理论假设，并进行数据分析。 研究结果表明，信息质量与响应性是学术用户持续使用的核心驱动因素，具体表现为生成内容的准确性与数据更新及时性显著影响用户感知有用性。有形性、伦理安全与同理心对持续使用意愿的直接影响较弱。有形性和同理心对感知有用性未达显著水平，伦理安全对期望确认的影响系数较低。究其原因，学术用户对AIGC的评估呈现工具属性优先的态势，用户更关注内容生成质量，而非界面美观、情感关怀等非核心需求，而伦理风险多为潜在风险，用户短期使用中难以直接体现，因而对期望确认程度影响较小。而期望确认通过提升感知有用性与用户满意度形成双重中介路径，揭示了学术用户从技术接受到持续依赖的动态机制。 最后，本研究以期望确认模型为理论基础，结合服务质量理论，针对AIGC在学术科研领域的应用，构建了AIGC服务质量的多维评价体系，并深入探讨AIGC服务质量对学术用户持续使用意愿的影响机制，通过实证分析验证相关假设，为AIGC服务质量与用户行为关系的研究提供了理论贡献，并为AIGC平台的优化与改进提供了实践指导，为“科研专用智能体”设计理论提供了新视角。
英文摘要：	The rapid development of Artificial Intelligence Generated Content (AIGC) technology is profoundly transforming the paradigm of academic research. With breakthroughs in deep learning, natural language processing, and big data technology, AIGC has evolved from a simple text generation tool into an intelligent partner capable of assisting with literature reviews, experimental designs, data analysis, and even academic writing. However, despite the increasingly widespread application of AIGC in academic settings, there is a lack of systematic research on how its service quality affects the continuous usage intention of academic users. As a result, AIGC developers lack corresponding theoretical support when optimizing their products, and academic users also face issues such as insufficient trust and low stickiness during the usage process. Therefore, exploring the mechanism of action between AIGC service quality and academic users' continuous usage intention is not only an inevitable requirement for theoretical deepening but also an urgent need for practical optimization. This paper takes the Expectation Confirmation Model as the core theoretical framework and combines it with the service quality theory to construct an AIGC service quality evaluation system that includes five dimensions: tangibility, information quality, responsiveness, empathy, and ethical security. The research is carried out in three stages: First, based on a literature review, an initial model is refined, covering 10 service quality dimensions and 30 specific indicators; Second, a semi-structured interview method is used to conduct in-depth research on 10 experts in the AIGC field and academic users who have long used AIGC. Combining qualitative analysis, the model is optimized to form a revised framework with five dimensions (tangibility, information quality, responsiveness, empathy, ethical security) and a total of 19 measurement indicators; Then, 265 valid data are collected through a questionnaire survey, and the structural equation model is used to verify the theoretical hypotheses and conduct data analysis. The research results show that information quality and responsiveness are the core driving factors for academic users' continuous usage, specifically manifested in the fact that the accuracy of the generated content and the timeliness of data updates significantly affect users' perceived usefulness. Tangibility, ethical security, and empathy have a relatively weak direct impact on the intention of continuous usage. Tangibility and empathy do not reach a significant level in terms of perceived usefulness, and the impact coefficient of ethical security on expectation confirmation is low. The reason is that academic users evaluate AIGC with a priority on its tool attributes. Users pay more attention to the quality of content generation rather than non-core needs such as interface aesthetics and emotional care. Moreover, ethical risks are mostly potential risks, which are difficult to be directly reflected in the short-term usage of users, thus having a small impact on the degree of expectation confirmation. Expectation confirmation forms a dual mediating path by enhancing perceived usefulness and user satisfaction, revealing the dynamic mechanism of academic users' transition from technology acceptance to continuous dependence. Finally, based on the Expectation Confirmation Model and combined with the service quality theory, this study constructs a multi-dimensional evaluation system for AIGC service quality in the context of AIGC applications in the field of academic research. It deeply explores the influence mechanism of AIGC service quality on academic users' continuous usage intention, verifies relevant hypotheses through empirical analysis, provides theoretical contributions to the research on the relationship between AIGC service quality and user behavior, offers practical guidance for the optimization and improvement of AIGC platforms, and provides a new perspective for the design theory of "intelligent agents dedicated to scientific research."
查看全文：	预览  下载(下载需要进行登录)