Formalization Processing Research Of Chinese Quantifiers And Group Relations In Domain Of Manufacture

With the development of mechatronics technology, multi-disciplines cross very tightly, what is increasing the complicities of manufacture systems. Hence, intellectualized mechatronics system designing is the tendency in future and Expert System which is the represent of Knowledge Engineering will exert more and more effects. Chinese Understanding technique was applied to knowledge acquisition for domain Expert System in this paper by formalizing the understanding results extracted from the analysis of design experiences, control experiences or knowledge got from experts describing in natural language to the knowledge for knowledge base of computers for providing Expert Systems an automatic knowledge acquisition interface.Upon the analysis of the characteristics of natural language in manufacture domain and the balance of the advantages and disadvantages of two current methods to NLU, the method based on knowledge was chosen as the main technique to understand domain Chinese. Firstly, according to the demand of Expert Systems' knowledge acquisition, several existing knowledge representation methods were analyzed and the mixed expression of Rule and Frame was applied to representation of the resource intension and extension. Secondly, upon the emphasized analyze of common quantifiers in the domain, a suitable class of quantifiers was given and the signification of the implied quantitative information to the sentence formalization processing was putted forward. Then, base on Resource Model, by the relations of resource intension and extension, group semantic information in Modern Chinese was analyzed and the limitation of separately depending on the quantifier semantic understanding was basically overcome. Based on the theory analyze, the elementary implementation of quantifiers and its structures, implied quantitative information and group relations understanding was introduced. Finally, the system was applied to sensor knowledge acquisition system with satisfactory result after debugging.