Speech Technology In Educated System

Education based on the communication between teacher and student. Speech is the naturalist way.One intelligence education system based on Speech Analyses,Text-To-Speech, Speech Recognition technology, including system structure and how to produce it. Have carried on the detailed discussion to essential technology and the principle, and proposed the improvement and the suggestion on the speech teaching system.Grammar and semantic analysis are the foundation; understanding is the first step. Chinese has its own characteristic, It has certain difficulty in analyzes and understood.Through word cuts and semantic analysis, this article has obtained in the understanding information. So it is possible to develop an intelligent speech system in specialized domain.Moreover, this article rests on the reasoning principle to make inferential reasoning more intelligent. Forecast semantics and linguistic environment, improved the intelligence of forecast. All above improved recognition grammar rule for enhancing the rate of speech recognition accuracy.Text-To-Speech and the speech recognition is the method. If we had the accurate semantics, we can realize our speech system through Text-To-Speech and Speech Recognition.This article uses XML language in Text-To-Speech rhythm to optimize the result.Through creating the XML grammar rule dynamically, we improve the recognition in small area. By rests on the reasoning principle we can analyses the area of semantic and using in different position.This article creates the recognition grammar rule of next step by dynamic alternate. By include the intelligent analyses and forecast, this recognition grammar rule reduced the recognition scope more reasonable. So the rate of recognition enhances greatly.This article reminds some thing to pay attention in using XML in SAPI. For example, Label nesting use, how to use certain labels and matters needing attention.