A linguistic analysis of the problems of Facebook (see translation)‎ and Google (translate)‎ applications : a selected sample of corpora

الملخص EN

Machine Translation or (MT) is considered one of the recent innovations in technology in the field of translation studies (TS).

This paper is an attempt to redefine nowadays users’ methods of translating corpora from Arabic into English and vice versa.

MT is a computational activity in which the translating process is done by using a bilingual or multilingual data set.

Corpus analysis is one of the fastest-growing methodologies in contemporary linguistics.

Many institutions and individuals use MT to translate corpora.

They resort to computational translation applications offered by Facebook (see translation) and Google (translate) to render their corpora (words, utterances, statements, speech, texts..,etc.

Regardless of the fatal mistakes sometimes found in the results of MT (For example, those mistakes related to diacritical markers), no one can deny that MT systems are built on gigantic lexical banks and dictionaries.

This paper presents a linguistic analysis of the problems of MT namely Facebook (see translation) and Google (translate) applied to a limited sample of corpora.