智能教学系统NKI-Tutor的知识查询设计

Design of Knowledge Query for NKI-Tutor

NKI-Tutor是国家知识基础设施面向应用的智能教学系统，其功能是将NKI丰富的知识传授给用户。知识查询是NKI-Tutor给用户提供的一种查询式学习模式。该文研究如何通过NKI-Tutor的知识查询界面实现用户与智能教学系统的交互而获得所需知识，提出了基于自然语言理解的知识查询方法，用于构造和实现NKI-Tutor中用户查询知识的交互界面。该方法包括3方面内容：(1)分析用户的知识查询形式，采集用户查询句型模板；(2)根据用户查询的知识特征，把知识查询映射到概念-关系模型上；(3)通过智能分词和模糊匹配构造查询结果。

NKI-Tutor is an application system for national knowledge infrastructure (NKI), aiming at tutoring users how to acquire knowledge from the giant knowledge engineering system. It is important that the NKI-Tutor ought to provide variety communication interfaces. This paper proposes an interface, which is content-based query. This interface has three significant features: analyzing the different types of knowledge query, and collectioning the query sentence templet; mapping of knowledge queries into the CR model; and gaining the feedback of queries with intelligentied splitword and fuzzy matching.