|
Technology Issues
|
Electronic translation aids involving Indian languages are still at an early stage
of development. Such tools would include word processors, dictionaries, management
tools, term banks, online dictionaries, online and visual thesauri, word-finders,
translation memory softwares, e-translators (=machine translation softwares) translation
teaching softwares, glossaries, corpora, spell checkers, grammar checkers, electronic
dictionaries, technical glossaries, gramatical glossaries, online tools, cultural
vocabularies and various look up facilities. All of them, except the machine translation
package, support the translator but do not perform the actual translation task.
NTM can facilitate technological advancement in machine translation by assisting
in the training and development of human technical resources for some activities
and supporting others through collaboration and providing synergy. NTM can come
in to help in the following areas is association with bodies like C-DAC and Technology
Development for Indian Languages (TDIL) etc.
|
a.
|
Building up of the necessary infrastructure, especially digital tools like the thesauri,
bilingual dictionaries, software for translation memory etc which have immediate
application for more efficient and effective translation
|
b.
|
Creation and maintenance of lexical resources like e-dictionaries, wordnets, language
analysis and synthesis tools, concordancers, frequency analysers etc through multi-institute
collaboration over a long period: here NTM could provide a common platform for teamwork
through continuous interaction by holding meetings, on-line discussions etc.
|
c.
|
Working out the copyright issues for source texts and translations in clean digital
form. NTM can ensure that this wealth of digital material is maintained in a standard
format with DTDs.
|
d.
|
Production and maintenance of high quality parallel corpora with annotations and
alignments being generated under LDC-IL project. Such corpora can be treated with
machine learning techniques to obtain machine translation systems.
|
e.
|
Promotion of an interlingua based approach along the lines of the ‘Universal Networking
Language’ (UNL) initiated by the United Nations in 1996 involving 15 countries.
IIT, Bombay has already developed various tools, techniques and resources for English
and Indian language MT which could be generalised.
|
|
|