Studies On Standard Subclass Language Of English-Chinese Machine Translation System Based On Globish

The 21st century is the information era. With the recent frequent international communication, the variety of language becomes an obstacle. The appearance of Esperanto, as a language in common use originally, tried to realize the communication of all nations. However, it didn't be popularized since some reasons. The chief reason is lack of speakers。The France, Jean-Paul Nerrière, proposed Globish, which is based on English, seems to bring us the new hope for communication with other nations. Although we can translate by interpreters, the high expenses and inconvenience always make us shrink back at the sight. Machine Translation can help people realize the aided communication with each other. Machine Translation is a branch of natural language processing, and it also involves knowledge in other fields. As the finiteness of human beings'understanding in natural language and the narrow limitation of computers in processing natural language, Machine Translation becomes a problem.In such situation, this thesis presents the thought of Globish (standard subclass language)-based English-Chinese Machine Translation system in limiting field such as tourist, which aims to be use the platform of Machine Translation ,by use of Globish,the standard subclass language,solves the language obstacle and meet the satisfactions of communication and understanding.This thesis summarizes the basic theories of Machine Translation and the correlation knowledge of natural language processing, describes the creation of system corpus, and presents the system flowchart and translation algorithm.One task of this thesis is to establish three libraries, one is the bilingual dictionary containing 1500 Globish words, using the structure of fixed-length fields and second-level index, other is a part travelling glossary, and another is bilingual example sentence-pairs in travelling, stored by extensible markup language(XML). The other task is to design the system flowchart according to the characteristics of limiting travelling field and the features of Globish. When the translation can not be comprehended, the system will try to output the picture together with the target words. In the process of the translation, this thesis uses an improved example-based Machine Translation algorithm. Sentence similarity can be calculated according to extracting abstract words firstly, and then decide whether to go to the further translation basing on the matching result. It also imports the translation memory technology.Globish (standard subclass language)-based English-Chinese Machine Translation system is a realizable simplified system, and its complexity is more simple than other natural language translation system. It can be applied and embedded in application system such as electrical dictionary, moble phone, MP3, MP4etc, which has the wide application.