Polysemy Of Adjectives In Natural Language Understanding And Its Application In The Field Of Product Designing

This thesis applies natural language understanding to the needs of mechanical product design analysis, and focuses on the polysemy of adjectives. Finally, the systems succeed to understand and analyze the user's needs expressed in the form of natural language. Then, we convert the results of the analysis into the conceptual design requirements or design parameters to support the following design.This thesis identifies a variety of natural language understanding and knowledge representation, describes some of the semantic knowledge representation and analysis of their advantages and disadvantages. Firstly, we focus on the concept of subordination theory and clustering methods, and in-depth discussion of the concept of subordinate tree. Secondly, we give definition of the polysemy of Chinese adjectives and relevant discussion, and classify the Chinese adjectives according to the causes resulting to polysemy of an adjective. Thirdly, we discuss in detail three methods for polysemy of Chinese adjectives processing, which are based on the noun, based on some spectial structures and for the adjective as a verb, respectively. For these methods, we need to establish subordinate concept trees for the adjectives and nouns. In addation, we should employ the cluster analysis for the third method. In this thesis, we give some examples for the cases: adjective as an attribute and a predicate respectively for semantic analysis. Finally, the effectiveness of methods for polysemy of Chinese adjectives processing is confirmed via an example in the field of mechanical product design.