Using context-dependent interpolation to combine statistical language and translation models for interactive machine translation
Philippe Langlais, George Foster · 2000
This work is in the context of TRANSTYPE, a system that watches over the user as he or she types a translation and repeatedly suggests completions for the text already entered. The user may either accept, modify, or ignore these suggestions. The system's proposals are selected and scored using a linear combination of a trigram language model and a translation model. We investigate the issue of how weights should be assigned to these two models in different contexts. 1 Introduction TRANSTYPE is part of a project set up to explore an appealing solution to the problem of using Interactive Machine Translation (IMT) as a tool for professional or other highly-skilled translators. IMT first appeared as part of Kay's MIND system (Kay, 1973), where the user's role was to help the computer analyse the source text by answering questions about word sense, ellipsis, phrasal attachments, etc. Most later work on IMT, eg (Blanchon, 1991; Brown and Nirenburg, 1990; Maruyama and Watanabe, 1990; Whitel...