J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. The automatic component of the LINGSTAT machine-aided translation system. In , 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....we have not spent enough time thinking about how to acquire, in as automatic a manner as possible, the lexicons and rule bases we use. Another interesting statistics based MT approach to using statistical techniques to generate MT systems is LINGSTAT developed by Dragon Systems, Inc. 230] [231], and [19] This work started in 1992 as a translation aid as a direct substitution system with a simple, hand generated finite state grammar for Japanese. The English glosses were based on bilingual dictionaries and the grammar was used to assign Japanese phrase attachment. This was quickly seen ....

J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. The Automatic Component of the LINGSTAT Machine-Aided Translation System. In ARPA Workshop on Machine Translation, Vienna, Virginia, March 1994.

....for statistical methods, has been addressed by techniques involving dynamic programming [151] and structural constraints [228] 230] Another interesting statistics based MT approach to using statistical techniques to generate MT systems is LINGSTAT developed by Dragon Systems, Inc. 19] 232] [233]. This work started in 1992 as a translation aid as a direct substitution system with a simple, hand generated finite state grammar for Japanese. The English glosses were based on bilingual dictionaries and the grammar was used to assign Japanese phrase attachment. This was quickly seen to be ....

J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. The Automatic Component of the LINGSTAT Machine-Aided Translation System. In ARPA Workshop on Machine Translation, Vienna, Virginia, March 1994.

....However, our use of statistics allowed us to avoid much of the traditional hand coding, and to produce a competitive MT system in nine months. Other statistical approaches to MT include CANDIDE [ Brown et al. 1993 ] which does not do a syntactic analysis of the source text, and LINGSTAT [ Yamron et al. 1994 ] which does probabilistic parsing. Both LINGSTAT and JAPANGLOSS require syntax because they translate between languages with radically different word orders. Our use of features in syntax, glossing, and semantics gives us the flexibility to correct translation errors, capture generalizations, ....

J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. The automatic component of the LINGSTAT machine-aided translation system. In Proc. ARPA Workshop on Human Language Technology, 1994.

....has great practical importance in Japanese English machine translation. Katakana phrases are the largest source of text phrases that do not appear in bilingual dictionaries or training corpora (a.k.a. not found words ) However, very little computational work has been done in this area; (Yamron et al. 1994) briefly mentions a patternmatching approach, while (Arbabi et al. 1994) discuss a hybrid neural net expert system approach to (forward) transliteration. The information losing aspect of transliteration makes it hard to invert. Here are some problem instances, taken from actual newspaper ....

J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. 1994. The automatic component of the LINGSTAT machine-aided translation system.

.... and Mitamura, 1992] ffl Abandoning automatic MT, and building software to assist human translators instead [Isabelle et al. 1993; Dagan and Church, 1994; Macklovitch, 1994] ffl Developing automatic knowledge acquisition techniques for improving general purpose MT [Brown et al. 1993b; Yamron et al. 1994; Knight et al. 1995] There have been exciting recent developments along all these lines. I will concentrate on the third thrust improving MT quality through automatic knowledge acquisition. Source Text (eg, Japanese) Target Text (eg, English) Target Text Semantics Source Text Semantics ....

....with a word skipping parser [Lavie, 1994; Yamada, 1996] that tries to find a maximal parsable set of words. Given reasonably accurate parsing systems (trained or handcrafted) it is possible to write transfer rules by hand and use a language model to do lexical and structural disambiguation [Yamron et al. 1994; Hatzivassiloglou and Knight, 1995] It is also possible to learn transfer rules from bilingual corpora automatically: both halves of the corpus are parsed, and learning operates over tree pairs rather than sentence pairs. A more ambitious, potentially powerful idea is to train directly on ....

Yamron, Jonathan, James Cant, Anne Demedts, Taiko Dietzel, and Yoshiko Ito. 1994. The automatic component of the LINGSTAT machine-aided translation system. In Proceedings of the ARPA Human Language Technology Workshop.

No context found.

J. Yamron, J. Cant, A. Demedts, T. Dietzel, and Y. Ito. The automatic component of the LINGSTAT machine-aided translation system. In , 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC