Interlingual Machine Translation
In a world with many languages there is a need to remove the language barrier to allow people to interact fully. Many techniques have been applied to form unique language translation machines. The machines will cater to the need of translating the different languages of the world in the quickest way possible. Without these machines each community in the word would remain isolated from the other, due to the language barrier.
Interlingual Machine Translation is one approach to machine translation within the rule-based paradigm. The interlingual approach is an alternative to the direct approach and the transfer approach.
In Interlingual Machine Translation the text to be translated in the source language is changed into an Interlingua. This Interlingua is a summarized, abstract and mathemetical language edition. It is able to describe all of the characteristics of all of the languages which are to be translated, instead of simply translating from one language to another. The desired language in the translation is then generated from the Interlingua.
Sometimes two interlinguas are used in translation. In that case the first interlingua covers more of the characteristics of the source language, and the other one has more of the characteristics of the target language. In this translation process the sentences from the first language are translated into the target language through two intermediate stages. The system may also be set up such that the second interlingua uses a vocabulary that is more specific to a certain domain, as this could improve the translation quality.
The above-mentioned system is based on linguistic proximity. A text in one language is analyzed once and then translated to many other languages with a structural similarity. This system is meant to improve the translation quality and is also used in pivot machine translation. But in pivot machine translation a natural language is used as a “bridge” between two more distant languages, instead of an interlingua. This is the case in translating to English from Ukrainian using Russian as an intermediate language, for example.
The interlingual approach to machine translation has advantages and disadvantages. The advantage of the interlingual method is that it requires few components to ‘learn’ translating from each source language to each target language. When adding a new language, just being able to translate from the new language to the Interlingua (and the other way around) is enough. And it also allows the analyzer and producer to be written by monolingual system developers. They only need knowledge of the language that is to be added and of the Interlingua. Another advantage is that it supports paraphrases of the input in the original language. And on top of that, it can deal with languages that are totally different from each other.
The disadvantage of this method is that the proper definition of an Interlingua is difficult. It is possible to define one for a specific domain, but it might well be impossible to come up with a unique definition for a wider domain. The ideal context for interlingual machine translation is thus multilingual machine translation in a very specific domain.
In the evaluation of machine translation, the machine translation systems and productivity can be assessed using numerous ways. One of them is how well does the translation fit the intended use. The other one is the nature of the whole translation process and the characteristics of the machine translation software. The translation must suit the culture and language patterns used by people who will read the translated material. And the software to be used in the process should produce high quality translations that are very close to the original text.