Machine translation of Chinese
ثبت نشده
چکیده
To the Western mind the Chinese language, particularly the written language, has long been a source of wonder. Because the Westerner has known an alphabet from childhood he can scarcely believe that the strange ideographic characters of Chinese can convey the whole gamut of human thought, from philosophy to platitudes. Even without the added complication of ideographic writing, the Chinese language is so difficult that it seems most unlikely that many Americans or Europeans will ever learn it well. Machine translation of Chinese would seem to offer the only realistic hope of giving the West ready access to the manners, achievements and aspirations of a fourth of the human race. The Indo-Chinese group of nations, with a population of about 750 million, is currently publishing in newspapers, journals and books about three billion words a year. Much less than 1 per cent of this vast output is now being translated and republished in English, French or German (undoubtedly a larger -percentage is being translated into Russian), yet nearly all of it would be of great interest to one Western group or another. Automatic translation is needed because human -translators cannot handle the volume or hope to acquire the special vocabulary needed to make good translations in a wide variety of technical fields. Almost as soon as the electronic computer was developed some 20 years ago, its potential usefulness for language translation was recognized. Before this potential could be exploited, however, the entire technology of information processing had to be raised to its present high level. The basic problem is to find a match between natural languages and computer languages. The term "computer languages" means formal, coded instruction to a machine capable of making precisely formulated decisions. This need for stringent formalization goes far beyond the description of languages traditionally provided by linguistic scholars. As a result languages have had to be analyzed afresh by linguists willing to recognize the necessity of precise statements for a machine. In return the computer, with its capacity for subjecting an enormous amount of data to formal treatment, has served the linguist as a powerful tool for experimentation. Thus, aside from practical results, the attempt at automatic translation can hardly fail to yield new insights into the general properties of languages. Several years ago a number of groups in the U.S. began a major effort aimed at machine translation of Russian into English. The work ranged from the highly theoretical to immediate practical efforts at computer programing. Today, under Government auspices, scientific and technical journals published in Russian are routinely translated by machine. The results, although far from perfect, have demonstrated that understandable and useful translations can be made automatically. Work on the machine translation of Chinese was undertaken in 1960 at the University of Washington. Research programs have also been initiated at the University of California at Berkeley, Ohio State University and elsewhere. Over this same period Soviet linguists have been hard at work on machine translation of Chinese into Russian. This article will describe the work of the Chinese-to-English machine translation program that the International Business Machines Corporation undertook in 1960 under a contract with the Air Force. The Chinese program has benefited greatly from ideas and machine techniques developed for an earlier Russian-to-English program. Perhaps the most important feature of that systern is a photographic "memory" containing hundreds of thousands of dictionary-like entries, any one of which can be found in a twenty-thousandth of a second [see top illustration on opposite page]. The need for such a large memory arose from an early recognition that a translation program could not be constructed simply by mechanizing existing grammars. It proved impossible to find any fabric of grammatical and syntactical rules that could be reduced to a manageable set of machine instructions. All spoken languages consist of a very large—in fact an infinite—set of conventions. In contrast, the operation of even the most complex machine can be expressed in terms of a relatively small set. In mechanical translation an order must be found in the larger set that makes it amenable to processing by MODERN GUIDED-MISSILE ALREADY POSSIBLE CARRY WITH WAR HEAD OF HYDROGEN BOMB AND ATOMIC BOMB, THEREFORE IT IS ONE KIND WEAPON WITH VERY BIG POWER OF DESTRUCTION.
منابع مشابه
The Use of Second-Person Reference in Advertisement Translation with Reference to Translation between Chinese and English
This research aimed to review the use of second-person reference in advertisement translation, work out the general rules, and provide guidance to translators. Using second-person reference is common in the advertising discourse. Addressing audiences directly involves their attention and in this way enhances their memorization of the advertised message. Second-person reference can be realized v...
متن کاملWord, Subword or Character? An Empirical Study of Granularity in Chinese-English NMT
Neural machine translation (NMT), a new approach to machine translation, has been proved to outperform conventional statistical machine translation (SMT) across a variety of language pairs. Translation is an open-vocabulary problem, but most existing NMT systems operate with a fixed vocabulary, which causes the incapability of translating rare words. This problem can be alleviated by using diff...
متن کاملSystem Description: Dependency-based Pre-ordering for Japanese-Chinese Machine Translation
This paper describes the Beijing Jiaotong University Japanese-Chinese machine translation system which participated in the 1st Workshop on Asian Translation (WAT 2014). We propose a preordering approach based on dependency parsing for Japanese-Chinese statistical machine translation (SMT). Our system achieves a BLEU of 24.12 and a RIBES of 79.48 on the Japanese-Chinese translation task in the o...
متن کاملChinese-to-Japanese Patent Machine Translation based on Syntactic Pre-ordering for WAT 2016
This paper presents our Chinese-to-Japanese patent machine translation system for WAT 2016 (Group ID: ntt) that uses syntactic pre-ordering over Chinese dependency structures. Chinese words are reordered by a learning-to-rank model based on pairwise classification to obtain word order close to Japanese. In this year’s system, two different machine translation methods are compared: traditional p...
متن کاملSYSTRAN Chinese-English and English-Chinese Hybrid Machine Translation Systems for CWMT2011
This report describes SYSTRAN’s Chinese-English and English-Chinese machine translation systems that participated in the CWMT 2011 machine translation evaluation tasks. The base systems are SYSTRAN rulebased machine translation systems, augmented with various statistical techniques. Based on the translations of the rule-based systems, we performed statistical post-editing with the provided bili...
متن کاملA Comparative Study of English-Persian Translation of Neural Google Translation
Many studies abroad have focused on neural machine translation and almost all concluded that this method was much closer to humanistic translation than machine translation. Therefore, this paper aimed at investigating whether neural machine translation was more acceptable in English-Persian translation in comparison with machine translation. Hence, two types of text were chosen to be translated...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012