|
Both extremes of machine and human translation are easily discarded as
valid models for contemporary translation.
Machine-Fully Automatic High Quality Translation
The idea of an independently acting, error-free translating machine is
as unrealistic as that of a human with pencil and paper. After spending
a great deal of money on the development of Machine Translation systems,
researchers gave up the idea of Fully Automatic High Quality Translation.
They have instead focused on using Machine Translation for the translation
of texts covering a highly restricted encyclopaedic and terminological
field and consisting of very simple sentence structures.
Humans & computers cooperate:
In the past five to ten years the language industry has spent enormous
amounts of money creating and marketing products that are able to convert
the total translation process in one go.
Products designed to automate the process of translation almost completely
are translation memories, software localization tools and machine translation
systems.
Translation Memories
A multilingual text archive containing (segmented, aligned, parsed and
classified ) multilingual texts, allowing storage and retrieval of aligned
multilingual text segments against various search conditions. (as defined
by: Expert Advisory group on Language Engineering standards. )
Examples:
IBM's Translation Manager
Atril's Déja Vu
Trados's Translator's Workbench.
|
