A research perspective on how to democratize machine translation and translation aids aiming at high quality final output

MT Summit VII Sept. 1999 A research perspective on how to democratize machine translation and translation aids aiming at high quality final output Christian BOITET GETA, CLIPS, IMAG-campus, F-38041 Grenoble cedex 9, France Christian.Boitet@imag.fr Abstract Machine Translation (MT) systems and Translation Aids (TA) aiming at cost- effective high quality final translation are not yet usable by small firms, departments and individuals, and handle only a few lan- guages and language pairs. This is due to a variety of reasons, some of them not fre- quently mentioned. But commercial, tech- nical and cultural reasons make it mandato- ry to find ways to democratize MT and TA. This goal could be attained by: (1) gi- ving users, free of charge, TA client tools and server resources in exchange for the permission to store and refine on the server linguistic resources produced while using TA; (2) establishing a synergy between MT and TA, in particular by using them jointly in translation projects where trans- lators codevelop the lexical resources spe- cific to MT; (3) renouncing the illusion of fully automatic general purpose high quali- ty MT (FAHQMT) and go for semi- automaticity (SAHQMT), where user par- ticipation, made possible by recent techni- cal network-oriented advances, is used to solve ambiguities otherwise computation- nally unsolvable due to the impossibility, intractability or cost of accessing the ne- cessary knowledge; (4) adopting a hybrid (symbolic & numerical) and "pivot" ap- proach for MT, where pivot lexemes arc UNL or UNL inspired English-oriented de- notations of (sets of) interlingual accep- tions or word/term senses, and the rest of the representation of utterances is either fully abstract and interlingual as in UNL, or, less ambitiously but more realistically, obtained by adding to an abstract English multilevel structure features underspecified in English but essential for other langua- ges, including minority languages. Keywords Machine Translation, Translation Aids, MT/TA synergy, Memory-Based TA, Ge- neralized Lexical Contribution, mixed MT approaches, SAHQMT Introduction Why speak about democratizing Machine Translation (MT) and Translation Aids (TA) when so many MT sys- tems and computerized lexicons are available off the shelf at cheap prices? Because what is on the market is adequate for "MT for watchers", or "informative MT" (see [8] for this terminology), or "dictionary aids", but inadequate when it comes to quality MT output or really efficient TA, necessarily based on large translation memories. More precisely, Machine Translation (MT) systems and Translation Aids (TA) aiming at cost-effective high quali- ty final translation are not yet usable by small firms, departments and individuals, and handle only a few lan- guages and language pairs. This is due to a variety of reasons, some of them not frequently mentioned, which are briefly analyzed in sec- tion 1, in which we also detail reasons for democratizing MT and TA. The following sections (2—4) explain in more detail the methods we propose to combine to attain this goal. 1 Why quality MT and TA are only for niches and rich users 1.1 Machine Translation (MT) Quality MT is almost …

Content truncated to 500 words