lower content
Flash Header
Machine Translation (MT)
What is Machine Translation?
Machine translation (MT) is the automated translation of one human language (natural language) to another through computers.

A typical machine translation process may involve numerous processes such as parsing, semantic and stochastic analysis, language modelling, corpus alignment, disambiguation, etc.

Sophisticated machine translation software adopts different strategies to "understand" and "transfer" the source language into the target language, rather than simply carrying out replacement and rearrangement of words. This is the most costly alternative and is often out of reach of most home users.
Will translation software (machine translation) one day replace human translators?
The richness and complexity of the human language is such that machine translation can never completely replace human translators. The breadth of human knowledge, the ambiguity in human interactions and the continual evolution of language mean that professional human translators will always be required.

Current MT applications are mainly utilized to help users assimilate information, for example, to understand the gist of a paragraph or a web page, or the preliminary screening of large amounts of information to identify sections that need to be translated more precisely by professional human translators.

Very seldom are MT applications used by users to disseminate information. Only a good human translator can accurately and precisely convey your message and all its nuances in an appropriate style to the target recipient.

Many in computer science consider a perfect MT system to be the Holy Grail!
If MT cannot offer the reliability or quality of a human translator, does this mean that MT has no practical use?
There are a number of systems available today that can produce output which is of sufficient quality to be useful. We all get by with a certain amount of human-generated poor quality content or translations every day. For many purposes, it is sufficient that the user receive as good an idea as possible of the meaning ("gist") of what he/she is reading.

MT offers two main advantages: speed and cost-effectiveness. The basic Dr.eye Translation Platform can easily translate your documents, web pages and text, a task that will take an army of translators to handle. By automating the translation process, MT also greatly lowers the incremental cost of each subsequent translation.

In this era of globalization, it is probable that in many domains, MT may be the only feasible solution for handling the sheer volume of time-sensitive cross-language interaction required.
How can the quality of MT output be improved to a usable level?
Accuracy rates for even the best translation engines are low for general-purpose usage. Reading raw machine translation output is not unlike trying to understand what a non-native speaker with a weak command of your language is trying to say to you.

One method of improving the quality of MT output is "tuning". In an ideal situation, linguistic analysis can be carried out on for a specific usage and customer, and specialized glossaries and language rules can be developed to substantially enhance the accuracy rate. Such customized translation solutions for use in restricted domains have been successfully deployed effectively around the world.
Jasa Pembuatan Website By IKT