Analysis And Research Of Structural Ambiguity For Natural Language Understanding In Intelligent Mechanical CAD

In intelligent Computer Aided Designed (CAD), the requirement analysis formarkets and customers, the reasoning and decision to designed projects need a mass ofinformation and knowledge, which are described basically in natural language. SoNatural Language Understanding (NLU) is the necessary for intelligent CAD systems.But the commonness of all natural languages is ambiguity. When understanding naturallanguage the most difficulty is disambiguation. Firstly this paper derives ambiguityfrom the difficulty of natural language, and analyzes the definition of ambiguity and theexisting reason of ambiguity and classifies in theory. Then this paper decides theemphasis is structural ambiguity and shows the strategy of disambiguation. The researchof theory serves for application. So the following three chapters show the conceptualtreatment, the choice of knowledge expressing, the disambiguation of structuralambiguity based on static knowledge base and the application in the mechanicaltransmission design.