Machine translation in Malayalam
What is Machine
translation? It is the translation of a language to another language without any human’s help. Machine
translation is new in India compared to Europe. An ancient tradition is human translation. From Indian
languages, many works like science, philosophy, folklore, religion etc are translated. To support translation,
machine translation system requires software for translation and automated dictionaries.
Machine translation is an important technology. Between English and any regional language of India, there is a
big market for translation. Examples of manual translation are from
English to local languages, translation of news is necessary. The
second example is in English, Hindi and in local language, translation of reports of government departments and
public sector units are necessary.
Have you heard of bilingual system? It is a translation system that produces translation between two languages.
What about multilingual systems? Here, for any given pair of languages, translations are produced. Multilingual
systems may be uni-directional or bi-directional.
There are 18 regional languages in India but still for administration, business purposes etc, the English language
is continued as a de-facto link. For effective communication and understanding, all information related with
agriculture, health, land, education etc should be translated to the regional languages. Like roads, transportation
etc, translation is also important.
Some of the
machine translation project in India includes MaTra, Anusaaraka, UNL-based MT between English, Hindi and Marathi
etc. MaTra and Anusaaraka are projects of translation from English language to Hindi. As the name implies,
UNL-based MT is between English, Hindi and Marathi. An
international project of the United Nations University is the Universal Networking Language
(UNL). For all major human languages creating an interlingua is
its aim. The Indian participant in UNL is IIT Bombay.
Using the UNL
formalism, the work is going on MT systems between English, Hindi and Marathi. Using an ‘enconverter’, the
source language is converted into UNL and then it is converted into the target language.
In India, Machine
translation is in younger stage.
C-DAC started the MT revolution. C-DAC started with Natural Language Processing (NLP). Then C-DAC developed a
parser to parse Hindi, Gujarati, Sanskrit, German and English. IIT Kanpur,
University of Hyderabad, NCST Mumbai and CDAC Pune etc are prominent firms with such translations. By funding the
projects, the Technology Development in Indian Languages (TDIL) played an important role.
are going on in the field of Machine translation in Malayalam. The importance of Machine Translation has
realized by the Ministry of Information Technology. Two approaches for Machine Translation are rule-based and
statistical. As the name implies, the rule-based approach uses the knowledge of two languages and also with a
set of rules for translation. Up to some extent parsing is involved. A popular method of using aligned parallel
corpora is in statistical approach. That means, for the X-Y pair, two documents in X and Y are
needed. The website link namely
srprojects.wetpaint.com/page/Malayalam+Machine+Translation helps you to know more about it. Some of the sites related to machine translation are ashttp://language.worldofcomputing.net/machine-translation/machine-translation-overview.html