自然语言理解中概念分类的研究及其在产品设计中的应用

发布时间：2018-02-14 07:41

本文关键词： 自然语言理解概念分类概念从属树静态知识库产品设计　出处：《西安电子科技大学》2013年硕士论文　论文类型：学位论文

【摘要】：在自然语言理解中,知识库起着至关重要的作用。本文通过对概念分类的研究，构建静态知识库，并将其用到产品设计中。首先，介绍了自然语言理解以及概念分类的现状。其次，说明了自然语言理解的中主流的知识表示方法，并选择概念从属树作为基于知识表示方法的重要补充。然后，，具体介绍了概念内涵和外延、概念属性约束以及概念分类的本质。分析了属性约束间的关联，提出以概念内涵中属性约束作为分类标准，对概念进行分类的方法，并构建概念从属树。其中主要介绍对已有概念的自动分类并将新概念在概念从属树中的动态定位作为补充。最后，按照前面的方法，建立概念从属树，并介绍概念从属树管理系统的构建方法。将此方法应用到智能仪器产品设计领域，构建相关概念从属树，建立起该领域的专业知识库，为实现领域内产品设计中的自然语言理解提供基础，提高了智能仪器设计的效率。
[Abstract]:Knowledge base plays an important role in natural language understanding. This paper constructs a static knowledge base through the research of concept classification and applies it to product design. Firstly, the present situation of natural language understanding and concept classification is introduced. Secondly, the mainstream knowledge representation method in natural language understanding is explained, and the concept dependent tree is selected as an important supplement to the knowledge based representation method. This paper introduces the connotation and extension of concept, the essence of concept attribute constraint and concept classification, analyzes the relationship among attribute constraints, and puts forward a method of classifying concept by using attribute constraint in concept intension as classification standard. It mainly introduces the automatic classification of existing concepts and the dynamic positioning of new concepts in the concept dependent tree as a supplement. Finally, according to the previous method, the concept dependent tree is established. The construction method of conceptual subordinate tree management system is also introduced. The method is applied to the field of intelligent instrument product design, and the related concept subordinate tree is constructed, and the professional knowledge base in this field is established, which provides the foundation for realizing the natural language understanding in product design in the field. The efficiency of intelligent instrument design is improved.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：J524;H0

【参考文献】