一种面向自然语言交互接口语义分析的MNet方法研究

Research on MNet Method for Semantic Analysis of Natural Language Interaction Interface

自然语言接口的语义分析任务是将自然语言指令转化为某种目标系统的形式化描述逻辑,是对自然语言的形式化建模,但语义分析模型在可视化、并发性、复杂度、一致性、可变性等方面还不够完善.在吸收语义网、深度网络、依存分析等现有概念基础上,提出一种综合的意元网络MNet的语义分析方法.MNet由意元、内关系、外关系及特征属性构成,并通过层次化递归的方式进行定义,期待面向从短语、句子到篇章的整体语义空间描述.从元关系、树结构和网结构3个过程设计了MNet一般构造算法.重点提出了自底向上规约式的MNet语义依存树构造算法,实验表明MNet能有效解决语义依存分析和自然语言控制接口问题.将SCADA系统的自然语言操控接口的语义分析过程转化成MNet的一般构造过程,为自然语言语义分析提供了一定的思路.

The task of semantic analysis of natural language interface is to transform natural language instructions into formal description logic of a certain target system,which is the formal modeling of natural language.However,the semantic analysis model is not perfect in the aspects of visualization,concurrency,complexity,consistency and variability.Based on the existing concepts of semantic web,deep network and dependency analysis,a comprehensive semantic analysis method of MNet is proposed.MNet is composed of meaning,inner relation,outer relation and feature attribute,and is defined by hierarchical recursion,which is oriented to the whole semantic space description from phrase,sentence to text.The general construction algorithm of MNet is designed from the three processes of meta-relation,tree structure and network structure.A MNet semantic dependency tree construction algorithm based on bottom-up specification is put forward.Experiments show that the algorithm can effectively solve the problems of semantic dependency analysis and natural language control interface.The semantic analysis process of specific problems such as natural language control interface of SCADA system is transformed into the general construction process of MNet,which provides a certain idea for natural language semantic analysis.