基于个性化知识地图的交互式信息检索系统研究—从用户认知的角度

In the era of knowledge economy, information requirement of users is diversified, personalized, and knowledge oriented, which challenges the traditional system-oriented information retrieval system. The proposed user-oriented project aims to explore the interactive information retrieval system based on personalized knowledge map from the perspective of knowledge comprehension and cognitive science. There are four main sections of the study. (1) Investigating cognitive behavior and requirement of users during the information retrieval process.　Cognitive factors and variables affecting information retrieval are detected and identified through observing and analyzing cognitive behavior of users. (2) Constructing user-oriented framework of interactive information retrieval system. A multiple presentation of information space and cognition space is built for the information retrieval system. (3) Mining and associating hidden concepts in information resource space. Information resource space is modeled through formal concept analysis, knowledge structure and relativeness is revealed and represented on a conceptual, semantic, and formal basis. (4) Constructing personalized knowledge map and presenting information retrieval results. The personalized knowledge map is built and presented based on thresholds setting by users, which suggests users with associated terms and presents information retrieval results in a form of semantic network. The planned study not only develops and improves the current information retrieval theories, but also has practical implications such as improving quality in knowledge services and promoting knowledge acquisitions and usability.

在知识经济时代，用户的信息需求呈现出多样化、个性化和知识化的趋势。这使得传统的以系统为中心的信息检索模型面临极大的挑战。本课题以用户为中心，从知识理解和认知科学角度出发研究基于个性化知识地图的交互式信息检索模型。研究内容包括（1）用户检索过程的认知行为及需求调查：通过观察和分析用户的认知过程，发现和识别影响用户检索的认知因素和变量；（2）面向用户的交互式信息检索框架模型：对检索系统的信息空间和用户的认知空间创建多元表示；（3）信息资源空间的隐含概念挖掘与关联：通过形式概念分析方法对信息资源空间进行数学建模，概念化、语义化、形式化地呈现和揭示知识结构和关联；(4)个性化知识地图构建与检索结果展示：根据用户设定的阈值来显示个性化的知识地图，提示关联词汇并以语义关联网络形式展现检索结果。本研究发展和完善了现有的信息检索理论体系，在促进知识获取与利用、提高知识服务质量等方面具有实践意义。

在知识经济时代，用户的信息需求呈现出多样化、个性化和知识化的趋势。这使得传统的以系统为中心的信息检索模型面临极大的挑战。本课题以用户为中心，从知识理解和认知科学角度出发研究基于个性化知识地图的交互式信息检索模型。本课题主要创新之处在于以下三方面。第一，围绕促进用户与系统的双向认知为研究目标，提出了一个以动态、交互、可适应性知识地图为核心的信息检索模型，该模型是系统导向与认知导向的信息检索模型的有机结合。第二，提出了基于概念格的知识地图构建方案，可视化地展现信息资源隐藏的概念及其关系，辅助用户对信息空间进行探索。第三，将用户交互行为和相关反馈相结合构建用户模型，更精准地识别用户兴趣，提高个性化检索的性能。与传统的信息检索系统相比，基于知识地图的交互式信息检索系统变被动的知识查找为积极知识探索，协助并促进用户认知学习过程中的同化和顺应，帮助用户构建完善的认知结构，最终使用户需求得到真正满足。