基于智能手机的维吾尔语语音控制系统的开发  

DEVELOPMENT OF UYGHUR VOICE CONTROL SYSTEM BASED ON SMART PHONE

作　　者：米尔阿迪力江·麦麦提 吾守尔·斯拉木[1,2] 努尔麦麦提·尤鲁瓦斯[1,2] 热依曼·吐尔逊[1,2] 艾尼宛尔·托乎提[2] Miradeljan Mamat;Wushour Ialam;Nurmamat Yolwas;Rayima Tursun;Anwar Tohti(College of Information Science and Engineering,Xinjiang University, IJrumqi 830046 ,Xinjiang, China;Key Laboratory of Xinjiang Multilingual IT, Xinjiang University, Urumqi 830046,Xinjiang, China)

机构地区：[1]新疆大学信息科学与工程学院,新疆乌鲁木齐830046 [2]新疆大学新疆多语种信息技术重点实验室,新疆乌鲁木齐830046

出　　处：《计算机应用与软件》2016年第6期220-223,305,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(60762006);国家工信部电子发展重大项目(159018);新疆自治区自然科技项目(2011211A012)

摘　　要：以实现维吾尔语命令词识别为目的,重点研究维吾尔语命令词识别系统在Android平台下的开发与实现过程,介绍系统开发难点、核心技术及系统典型的几个功能。系统主要由Android开发包、Eclipse集成开发环境和API接口进行开发,并且通过自动选型规则来实现维汉英多种文字的正确显示及处理等问题,针对广大用户的不同说话方式,重新构建维吾尔语语音语法文件,解决各地不同方言问题。在一般实验室环境下做实验得到了90.56%的正确识别率和85.00%的成功执行率等测试结果,表明维吾尔语非特定人命令词识别研究中语法文件的结构及构建对系统有不同的影响。With the purpose of implementing Uyghur command words recognition, we elaborately studied the development and implementation process of Uyghur command words recognition system on Android platform,introduced the development difficulties,core technologies and typical functions of the system.The system was developed mainly using Android SDK,eclipse integrated development environment and API interfaces,and realised the functions of correct display and processing of multiple texts of Uyghur,Chinese and English through automatic styles selection rule.Aiming at different speaking styles of the majority of users,we rebuilt Uyghur voice and grammar files,and solved the problem of different dialects around the Region.Moreover we gained the testing results of right recognition rate of 90.56% and the successful implementation rate of 85% in the experiment made in usual Lab condition,this showed that in the research of Uyghur non-specific command words recognition,the structure and construction of grammar files had different effects on system.

关 键 词：ANDROID平台 维吾尔语 关键词识别 槽语法 命令词识别 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]