Qualifying The Verb Phrases And Noun Phrases In The Field Of Semantic Analysis

As an important branch of artificial intelligence, the development of the technique of Natural Language Processing (NLP) has become a bottleneck in recent years, but has some substantive breakthrough in restricted domain. This paper attempts to make semantic analysis on verb phrase and noun phrase of the education of billiards domain with the technique of NLP. Relatively speaking, because what we face is restricted surrounding (billiards surrounding), so the amount of vocabulary is not very big and the structures of phrase is not very hard. It only involves the vocabulary of billiards sport and some basic structures of phrase.First, some methods of expressing semantic knowledge are introduced, and Conceptual dependency theory is chosen as the foundation of the semantic analysis, through analysis and compares of several different semantic methods. The concept and the intension and the extension of it are expatiated. The relation of the concept of noun and some verb in this domain are denoted by setting up conceptual dependency trees. Second, the method of semantic analysis on verb and the method of semantic analysis based on event on verb phrase are brought forward by researching characteristics of verb phrase in billiards domain. And the templates of verb, static event and dynamic event with the method of knowledge expression based on frame are built up. Third, according to the characteristics of noun and noun phrase in billiards domain, a classification to noun and noun phrase is made, the templates of noun, dynamic noun and noun phrase are built up, and some different methods of semantic analysis are brought forward according to different classes of noun phrase. Finally, the NLP system in restricted domain is built up, which is used for actualizing the understanding of verb phrase, noun phrase and simple sentence in this domain, The system runs well after debug.