联系我们

论文查询结果	返回搜索

论文编号：	16060
作者编号：	2120243831
上传时间：	2026/6/4 18:03:25
中文题目：	多智能体AI产品的可视化协作模式与任务目标导向对用户持续使用意愿的影响研究
英文题目：	Research on the Influence of Multi-Agent AI Products’ Visualized Collaboration Modes and Task Goal Orientation on Users’ Continuance Usage Intention
指导老师：	任星耀教授
中文关键字：	多智能体AI产品；可视化协作模式；任务目标导向；感知任务执行效率；认知负荷
英文关键字：	Multi-Agent AI Products; Visualized Collaboration Modes; Task Goal Orientation; Perceived Task Performance Efficiency; Cognitive Load
中文摘要：	随着人工智能技术的快速发展，面向C端用户的多智能体AI产品得以涌现。该类产品能够调用多个具有特定角色分工的智能体，以团队协作方式处理复杂任务，并将协作过程以可视化形式呈现给用户。尽管该类产品在功能上具备优势，但其市场表现却并不稳定，面临用户留存率偏低的问题。 用户是否愿意持续使用，是AI产品能否成功的关键。交互界面设计对用户持续使用意愿具有重要影响。多智能体AI产品的可视化协作模式是其交互界面设计的最大特点，现有研究多关注言语元素以及静态视觉元素对用户持续使用意愿的影响，缺乏对可视化协作模式这一动态视觉元素的探讨。此外，交互设计元素的效果还会受到多种情境因素的影响，用户在使用AI产品完成任务时，有些任务的核心目标是高效获得产出，而另一些任务的核心目标则是清晰地理解和掌握过程，现有研究缺乏对任务目标导向这一情境因素的探讨。因此，本文的研究问题是如何通过恰当的可视化协作模式设计，实现与任务目标导向的匹配，进而改善用户体验，提升用户对多智能体AI产品的持续使用意愿。 在文献综述的基础上，本文采用情景实验法检验假设。两项实验结果表明，可视化协作模式和任务目标导向对用户持续使用意愿的交互影响显著，即在结果导向任务中，并行（vs.串行）可视化协作模式通过提高用户的感知任务执行效率，对持续使用意愿产生正向影响；在过程导向任务中，串行（vs.并行）可视化协作模式通过降低用户的认知负荷，对持续使用意愿产生正向影响。 在理论上，本文从AI–AI协作视角出发，将可视化协作模式作为AI产品动态视觉元素研究的新变量，揭示了其与任务目标导向对用户持续使用意愿的交互影响。在实践上，本文为多智能体AI产品的交互界面设计提供了指导，有助于AI产品设计者结合任务目标导向，更好地设计可视化协作模式。
英文摘要：	With the rapid development of artificial intelligence, multi-agent AI products targeting C-end users have emerged. Such products can invoke multiple agents with specific roles to handle complex tasks through teamwork, presenting the collaboration process to users in a visualized form. Although these products possess functional advantages, their market performance remains unstable, facing the challenge of low user retention. Whether users are willing to continue using a product is the key to the success of AI products. Interface design exerts a significant influence on users’ continuance usage intention. The visualized collaboration modes of multi-agent AI products represent the most prominent feature of their interface design. However, existing research primarily focuses on the impact of verbal and static visual elements on continuance usage intention, lacking exploration of visualized collaboration modes as a dynamic visual element. Furthermore, the effectiveness of interaction design elements is influenced by various contextual factors. When users perform tasks with AI products, the core goal of some tasks is to obtain output efficiently, while for others, it is to clearly understand and master the process. Existing research lacks an investigation into the contextual factor of task goal orientation. Therefore, the research question of this thesis is how to achieve a match between visualized collaboration modes and task goal orientation through appropriate design, thereby improving user experience and enhancing users' continuance usage intention toward multi-agent AI products. Based on a literature review, this thesis employs the scenario-based experimental method to test hypotheses. Results from two experiments indicate that the interaction effect between visualized collaboration modes and task goal orientation on continuance usage intention is significant. Specifically, in result-oriented tasks, the parallel (vs. serial) visualized collaboration mode exerts a positive impact on continuance usage intention by increasing users’ perceived task performance efficiency. In process-oriented tasks, the serial (vs. parallel) visualized collaboration mode exerts a positive impact on continuance usage intention by reducing users’ cognitive load. Theoretically, from the perspective of collaboration between AI agents, this thesis introduces visualized collaboration modes as a new variable in the study of dynamic visual elements of AI products, revealing its interaction with task goal orientation on users’ continuance usage intention. Practically, this thesis provides guidance for the interface design of multi-agent AI products, helping designers of AI products integrate task goal orientation to better design visualized collaboration modes.
查看全文：	预览  下载(下载需要进行登录)