基于Hadoop与XMPP的云翻译系统的研究与实现  

Cloud Translation System Based on Hadoop and XMPP

作　　者：余正红[1] 伍永豪[1] 邓娟[1] 王俊[1] 

机构地区：[1]武汉科技大学城市学院,武汉430083

出　　处：《计算机与数字工程》2014年第2期239-242,260,共5页Computer & Digital Engineering

基　　金：湖北省教育厅科研项目(编号:B20121104)资助

摘　　要：针对传统机器翻译系统准确性差、人工翻译成本高等缺陷,提出了一种基于Hadoop云计算框架与XMPP协议的云翻译系统解决方案,结合传统翻译技术和Hadoop云计算框架,利用XMPP在异构系统之间的互通,建立用户、译员和对象的三方互助云平台。该系统可挖掘互助沟通过程中的庞杂的语料资源,具有语料库数据量大,翻译准确、翻译效率高、智能性强等特点,解决了人工翻译成本高、机器翻译歧义性大等问题,实现了不同语种人群通过互联网进行文字即时通信时的多语无障碍沟通。Traditional machine translation systems have disadvantages such as poor accuracy,labor cost and so on.This paper proposed translation system solutions based on cloud computing Hadoop framework and XMPP protocol,combining traditional translation technology and cloud computing Hadoop framework which used XMPP interoperability among heterogeneous systems,created a user,the interpreter and the object of the tripartite mutual cloud platform.The cloud translation system can be mutual communication process mining complex corpus of resources.It has several brilliant features,such as a large corpus of data,accurate translation,translation efficiency,intelligence and so on.By using this system we solved the high cost of manual translation,machine translation ambiguity and other issues,to achieve a different language groups for text instant messaging via the Internet when multilingual communication barrier.

关 键 词：XMPP Hadoop框架 云翻译系统 机器翻译 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]