联系我们

论文查询结果	返回搜索

论文编号：	16090
作者编号：	2120243885
上传时间：	2026/6/5 23:01:02
中文题目：	人智交互情境中的数据意义建构研究
英文题目：	A Study on the Construction of Data Sensemaking inHuman-Intelligence Interaction Contexts
指导老师：	李樵
中文关键字：	数据意义建构；生成式信息检索系统；社会特征；对话智能。
英文关键字：	data sense-making; generative information retrieval systems;social characteristics; conversational intelligence。
中文摘要：	数据驱动社会中，数据已成为重要的战略资源，但数据价值的实现不仅依赖数据本身，更取决于用户对数据的理解、应用与反思能力。随着生成式信息检索系统逐渐成为用户获取与解释信息的重要工具，数据意义建构过程也由个体独立完成转向人与系统协同完成。相比传统信息检索系统，生成式信息检索系统不仅能够整合信息并生成解释性内容，还在交互过程中呈现出一定的社会特征，如感知、适应与预测能力。这种社会化交互方式可能进一步影响用户的数据理解与认知判断。因此，探究生成式信息检索系统社会特征与用户数据意义建构之间的关系，对于深化人智交互研究、优化生成式系统设计以及提升用户数据素养具有重要意义。 本研究基于社会化理论、参与式意义建构理论与布鲁姆分类理论，从对话智能与社交智能两个维度刻画生成式信息检索系统的社会特征，并将数据意义建构划分为记忆、理解、运用、分析、创造与反思六个阶段。在研究方法上，采用实验法、回归分析法与内容分析法相结合的方式，以 30 名被试为对象，围绕交通数据任务开展人智交互实验，并结合用户自报告数据分析系统社会特征与不同层级数据意义建构之间的关系。研究进一步通过相关分析与回归分析，对系统社会特征内部关系及其影响机制进行了检验。 研究结果表明，生成式信息检索系统的社会特征对不同层级的数据意义建构具有差异化影响。低层级的数据意义建构对系统社会特征依赖较弱，而在创造与反思等高层级认知活动中，系统社会特征的作用更加明显。其中，社交智能有助于促进复杂情境中的数据理解与创造性使用，对话智能则能够提升用户的信息整合与持续交互体验。此外，用户在创造阶段更重视系统回应的情境适配性，在反思阶段则更关注系统能否提供不同观点与批判性反馈。本研究从数据意义建构层级视角出发，探讨了生成式信息检索系统社会特征在人智交互中的作用机制，丰富了数据意义建构与人智交互领域的相关研究，也为生成式信息检索系统在不同认知任务中的交互设计与社会智能优化提供了理论参考。图 7幅，表 76 个，参考文献 101 篇。
英文摘要：	In today’s data-driven society, data is increasingly regarded as a core strategic asset. However, the value of data does not emerge automatically from the data itself; rather, it depends heavily on users’ abilities to interpret, analyse and apply it in specific contexts. With the rapid development of generative information retrieval systems, users are no longer relying solely on their own cognitive processes to understand data. Instead, data interpretation is gradually becoming a collaborative activity between humans and intelligent systems. Unlike traditional retrieval tools that mainly provide links or isolated information, generative information retrieval systems are capable of producing integrated explanations, maintaining contextual dialogue and responding to users in a more socially interactive manner. Features such as conversational feedback, contextual awareness and adaptive responses make these systems appear increasingly “ social ” during interaction, which may further shape how users perceive and construct meaning from data. Against this background, investigating how the social characteristics of generative information retrieval systems influence users ’ data meaning construction is important for understanding human – AI interaction and improving the design of intelligent systems. Drawing on socialization theory, participatory meaning construction theory and Bloom ’ s taxonomy, this study examines the social characteristics of generative information retrieval systems from two dimensions: conversational intelligence and social intelligence. At the same time, data meaning construction is divided into six cognitive levels, including memorization, comprehension, application, analysis, creation and reflection. To explore these relationships, the research combines experimental methods, regression analysis and content analysis. Thirty participants took part in human–AI interaction experiments based on traffic-data tasks. User self-reported data, interview materials and interaction records were collected to analyse how system social characteristics relate to different levels of data meaning construction. Correlation and regression analyses were further conducted to identify the internal relationships among system social characteristics and their mechanisms of influence. The findings suggest that the social characteristics of generative information retrieval systems affect different levels of data meaning construction in different ways. At lower cognitive levels, users show relatively limited dependence on system social characteristics. However, in higher-level activities such as creation and reflection, these characteristics become more influential. In particular, social intelligence appears to support users in understanding and creatively applying data within complex situations, while conversational intelligence contributes to information integration and sustained interaction. The study also finds that users pay closer attention to the system’s contextual adaptability during the creation stage, whereas at the reflection stage they value the system’s ability to provide alternative perspectives and critical responses. By examining data meaning construction from a layered cognitive perspective, this study further explains how the social characteristics of generative information retrieval systems operate within human–AI interaction. The research contributes to studies of data meaning construction and human – AI collaboration, while also providing theoretical support for the optimisation of conversational design and social intelligence in generative information retrieval systems. 7 figures, 76 tables, 101 references.
查看全文：	预览  下载(下载需要进行登录)