The Research Of Word Segmentation Based On Statistics And Semantic Its Application In Production Design

This paper designs and realizes the participle system model based on statistic and semantic analysis , and its application in the product design process. This model in the whole field natural understanding system on the base of natural language form to describe the user needs into a computer can identify concept design requirements.This paper analyzes the characteristics of Chinese word segmentation and the existing problems and combine existing word segmentation method, put forward the mechanical participle corpus, statistical participle and the semantic analysis of combining participle model, and the participle embedded into the whole field model system of Chinese understanding. Mechanical participle stage, creatively proposed first positive maximal matching method, and then reduce word segmentation method draw all forms. Eliminate ambiguity stage, first using the corpora statistical methods once disambiguation, statistical its co-occurrence degree and tie-in, and the tie-in situation feedback to statistics in the dictionary, realize participle system and the natural language understanding system of self-improvement and benign interaction; Then the semantic analysis of partial formal structure and gerund structure, combining the concept of knowledge representation method, affiliate of segmentation results further disambiguation, improve the segmentation system accuracy. Based on the actual will this participle model is applied to product design of user demand analysis field.