摘要
语音识别技术是近年来信息领域重要的科技发展技术之一。作为一门交叉学科,语音识别正逐步成为信息技术中人机接口的关键技术。探讨了为数据库查询提供自然语言接口的可能性,介绍了图书资料查询系统中语音识别的基本框架,并详细描述了采用微软Speech SDK技术实现图书资料查询的详细过程。给出了查询流程,基于状态转换图的词法分析和语法分析方法,将疑问句转换成SQL查询语句的方法,以及由查询结果生成答句的方法。
Speech recognition is one of most important pattern recognizing technology in recent years. It combined several typical information technologies. Speech recognition is developed to an essential interface technology between man and machine. Have discussed possibility providing natural language interface for data base inquiry, process having introduced that books and reference materials inquires about the fundamental frame that pronunciation distinguishes in system, and having described that the Speech SDK technology realizes detail that books and reference materials inquires about detailedly adopt Microsoft. Have given inquiry out technological process, have inquired about the sentence method, and have generated the method replying a sentence from inquiring about result owing to that state changing lexical analysis and grammar analysis method pursuing, an interrogative sentence is changed becoming SQL.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第24期6423-6426,共4页
Computer Engineering and Design
基金
国家自然科学基金重点项目(69933030)
湖南省教育厅基金项目(06C440)
关键词
数据库查询
自然语言接口
词法分析
语法分析
语义分析
database query
natural language interface
lexical analysis
syntax analysis
semantic analysis