Domain NLU Discourse Model And Its Application On Requirement Analysis Of Mechanical Products Design

This text applies nature language understanding discourse analysis to the mechanical product design, through understanding and analyzing to the user's needs that are expressed in the form of natural language, and transform analysis result into the designing requirement of the concept or design parameter, offer a good design environment to user and designer.First of all, some current methods and theories of NLU knowledge expression are presented and compared, the advantages and disadvantages of each method are shown, and then a way to express NLU discourse information is brought and applied to the construction of NLU Discourse Knowledge Base. Second, according to the main problems of discourse analysis, the clustering analysis and association analysis method is presented as the kernel process to solve discourse problems, and then the principle of concept clustering analysis is talked and the method of noun-clustering and event-clustering are shown, next, the inhesion reason of association is mentioned and the implemention way is advanced. Third, the Domain NLU discourse model is designed and realized according to the model.At last, the discourse analysis is applied to the design of the Man-machine interface in requirement analysis of machine design, by testing the system, the model is satisfied.