One of the domains of translation is corpus- based translation studies. As the writer says it is a kind of strategy for translation. He, later on, explains the parallel corpora which comprises text in two or more languages. It means there is a monolingual corpora, there is a bilingual or bi directional corpora and so on. The writer, then, explains the comparable corpora. It is a corpora which includes text collections whose size and content may vary. The composition and size of the corpora and their representative ness is relevant to respective languages. The writer says this article is the result of a study which was the corpus- based study of a connotation of non- literary language. It scrutinized the meaning of color words as it exists in linguistic expressions like to see red, to fell blue, and green with envy. It also explains what factors are responsible for activating the connotative meanings of the color words.

The writer mainly focused on the fact that corpus data can be usefully exploited in order to identify series of translation equivalents for words, expressions, idioms, ... etc. Another important factor mentioned by the writer is the reference to corpora. In fact, reference makes it possible to identify where differences and similarities lie across languages. Some case studies are provided by the writer to clarify the use of corpora. To sum up, the corpora has different types. The main aspects of corpora are comparability and reference. Usually, the use of TL corpora is for the sake of natural translation. Another important factor is that the choice in translation is related to choice in source language and it is done with the help of comparing chosen expression against its possible alternative along with the paradigmatic axis. Finally, the writer says corpus is used very much and the word profiling is therefore a much more straight forward matter that it was a few years ago. He emphasized that the combination of automatic processing, manual analysis and greater awareness of how language makes meaning help translator to reach equivalence very easily.