首页 | 本学科首页   官方微博 | 高级检索  
     

语音识别技术在数字图书馆检索系统中的应用
引用本文:叶小榕,邵晴. 语音识别技术在数字图书馆检索系统中的应用[J]. 科技导报(北京), 2008, 26(18)
作者姓名:叶小榕  邵晴
作者单位:中国科学技术信息研究所,北京,100038;中国互联网络信息中心,北京,100190
摘    要:概要介绍了语音识别技术和微软语音开发应用程序接口(Microsoft Speech API,SAPI)。语音识别技术随着计算机技术的发展而逐渐成熟,使计算机能够识别用户的语音输入,记录下用户的语音信息并执行相应的命令。微软的SAPI是Windows操作系统下的语音识别开发平台,它开发快捷,有设计良好的运行机制、识别引擎和调用接口,能够模块化组件开发。其次,利用SAPI,设计实现了应用语音识别技术的数字图书馆检索系统。此系统使用户在使用数字图书馆时检索更加方便,读者说出想查询的内容,系统即可完成整个检索过程并显示查询结果。此系统基于SAPI平台开发,采用了MySQL作为后台数据库。此检索系统分为一站式检索和智能检索。其中,一站式检索为读者提供了通用语音识别框,读者检索时无需使用鼠标和键盘,只需说出自己想要查找内容的关键词即可查询。一站式检索系统不仅能够实现对图书馆数据库的关键词检索,而且在界面上还同步提示出可选关键词、现有馆藏书目、借阅情况等信息,协助用户检索信息。而智能检索是在一站式检索的基础上,结合了中文分词技术,进一步降低搜索难度方便用户使用。用户检索时无需考虑关键词、检索语法,只需像平时一样说出想检索的内容,检索系统会自动识别用户的语音,然后进行智能分词,再通过转换过程,过滤出关键词并生成专业的检索语法进行检索,并最终返回检索结果。整个识别、分词、转换过程由系统自动完成,无需读者干预。随着技术的进步,结合了语音识别的检索系统将使数字图书馆能够更加方便快捷地为读者服务。

关 键 词:数字图书馆  语音识别技术  微软语音识别API

Speech Recognition Technology in Retrieval System of the Digital Library
YE Xiaorong,SHAO Qing. Speech Recognition Technology in Retrieval System of the Digital Library[J]. Science & Technology Review, 2008, 26(18)
Authors:YE Xiaorong  SHAO Qing
Abstract:In the first part,this paper presents an overview of the speech recognition technology and the Microsoft Speech API(SAPI).The speech recognition technology enables the computer to identify the user's speech input,record what the user says and carry out the corresponding orders.The SAPI provides a speech recognition development platform.Then,a speech recognition retrieval system for the digital library is designed and implemented on the SAPI platform.It deals with the user's query process.The user simply says what he or she wants without using the mouse and keyboard and the system will complete the entire retrieval process automatically.It includes one-stop retrieval system and intelligent retrieval system.The one-stop retrieval system provides a general speech recognition input box,which not only identifies the which the user says,but also can prompt the optional keywords,library collections,and so on.The intelligent retrieval system,which is built on the basis of one-stop retrieval system,combines it with the words segmentation system.So the user just says what he or she wants to retrieve naturally,and without considering the keywords and the professional retrieval syntax.The system can automatically handle the whole recognition,segmentation and retrieval processes.Now,with the development of the technology,this speech recognition retrieval system of the digital library will provide more convenient services for all users.
Keywords:digital library  speech recognition technology  Microsoft Speech API
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号