subject: Drawbacks Of Web Based Translation [print this page] Day by day advanced web based translation models approaching to market and if used laboriously can be gainful. Translation done by web-based services provides easy, cheap and fruitful way to translate pile of documents in other languages. This might be appropriate while doing itself and can save labor to get translation by alternate way or it might give you a fair idea about doing translation from web based is authentic or not. The accuracy of computer generated translation is depending on various factors for example quality would be pleasant in translation from Hindi to Marathi due to resemblance of source and target language. Which could be unacceptable incase of Hindi to Telugu.
Furthermore, Web based translation is unable to do justice with text having uncertainty, confusion, double-entendre or any complexity of human language. Depending on the need of circumstances, translation should be precise and specific for example in case of legal document. Whereas, translation of technical or scientific text needed more attentiveness because scientific words are different from daily use words and linguistic errors in scientific or medical document could be ruinous.
There are languages in which computer based models prefer same translation for two words, it was not wrong but depend on specific condition, and it could be result into wrong translation. For example translation of separate and different is in Hindi. Now consider segment like Do not do it by different way just separate them from other translation done by web based model can create confusion here and could not give satisfactory result than it become essential to search a substitute action. Hindi translators are proficient to take care of this sensitive segment.
Additionally, There are lots of words which can become as a source of error in case of computer generated translation. For example Light it can be a use differently in English, can be used in mathematics or chemistry. Use of those words in different domain needed professional translator who can translate as per source without affecting the meaning as well as quality of translation. In such cases artificial intelligence is not capable to resolve linguist complications.
There is an intermediary language known as interlingua which is an imaginary state to understand properly the conclusion of source text. Interlingua adds two more steps in translation:-
1.Convert source text into intermediary language. (Understanding of source text)
2.Convert intermediary language into target language. (Sensible sentence formation)
Web based models are in sufficient to carry out above two step reliably.
There is always such word in every language which includes cultural context, human behavior and non-literal meaning which is beyond the range of web based models for translation. Whereas translation companies having such competent translators, who work in their native languages, can easily understand the demand of source text and fix linguistic issue. There is always place for computer generated translation but they are poor substitute for translation companies.
