Collection of dictionary data through Internet translation service

MT Summit VII Sept. 1999 Collection of Dictionary Data through Internet Translation Service Keisuke Nakayama Akira Kumano R&D Center, Toshiba Corporation R&D Center, Toshiba Corporation Japan Japan Abstract We have developed an Internet translation service, which we began to provide in 1997 for English to Japanese translation and in 1998 for Japanese to English. In this service, users send a translation request from a web page and receive by e-mail the result of the translation outputted by Toshiba’s machine translation system. As in other similar serv- ices, users can specify English-Japanese word pairs(dictionary data) when making a transla- tion request. What distinguishes our service from others is that our service system con- structs users’ own dictionaries on the server and helps them with this work by extracting words which the system expects to improve the system's translation quality if included in the dictionaries. With this function, users can efficiently add new word pairs so as to up- grade their own dictionaries when requesting re-translation. The dictionary data thus ob- tained from users can be utilized to improve the system dictionary on the server also. 1 Introduction Through the spread of the Internet, currently a wide variety of services, most typically, homepage infor- mation retrieval, is available on the Web. Here, peo- ple can easily access foreign sites just like domestic sites and thus have many more opportunities to en- counter foreign languages, predominantly English. As such, the need of machine translation which helps us- ers understand Web sites written in other languages is rapidly increasing. In Japan, numerous software products which translate English documents on the Internet directly into Japanese have been developed and are being sold on the market. This translation software is now used by a great many users on their personal computers. At the same time, several Internet translation services, in which users send their documents to the translation server through the Internet and obtain the translation result from the server, are also available. Our transla- tion service named “MT Avenue,” which we will de- scribe in this paper, is one example of such machine translation services. In the meantime, in terms of translation quality, the current machine translation systems are still in their development stage; therefore, expansion of dic- tionary knowledge, including that of newly coined words, is indispensable to achieve higher accuracy. We have devised a method to collect and accumulate dictionary knowledge, which forms the core knowl- edge of the whole translation system, and began an Internet translation service for English to Japanese and Japanese to English translations. One of the innovative features of this service is that it stores user dictionaries on the server, which are a compilation of translation pairs obtained by users when they send a translation request to the server. Furthermore, the system reduces users' task by ex- tracting candidate words from users' text and present- ing them along with the translation result. By refer- ring to this word list, uses can add new translation pairs and request re-translation. This …

Content truncated to 500 words