Machine translation output assessment and its impact on reading comprehension

Machine translation is now widely used everywhere; However, its role as a language learning tool has not been confirmed, as there are concerns about its quality. However, if we compare the machine translation output with the output produced ten years ago, we see a significant improvement in its quality, especially in terms of vocabulary and grammar. Machine translation can be defined as: the process by which, using electronic devices, input can be provided from one language and output delivered in another language. When machine translation became available on smartphones, it gained universal acceptance because of its benefits such as free and easy access. In the field of education, many learners use this technology every day for various personal as well as academic purposes. These goals mainly include understanding a text that is not written in the native language or translating different texts from different languages into other languages and delivering it as homework. Machine translation can help learners gain a quick understanding of a text written in a language other than their mother tongue by producing an incomplete version. The aim of this research was to assess the quality of machine translation and its impact on students’ reading comprehension.

Three types of texts were selected with varying levels of difficulty. These texts were translated once by a human translator and once by machine translation (Google Translator). Finally, six texts were obtained. The output of machine translation was evaluated and analyzed. Postgraduate students who happened to use machine translation more frequently were then randomly divided into six groups, each group reading one of these texts and answering multiple choice comprehension questions at the end of the text. The T-test was performed on the data and it was found that from the three types of texts, the two types of texts, despite having some lexical and grammatical problems, were able to compete with human translation.

The data showed that the quality of machine translation is improving and has now reached a degree of quality that can be used as a tool in educational environments. Some guidelines were also given on how to use this technology in the classroom.

This study attracts attention of language educators to MT and its use in language teaching. It suggests that language educators should be trained to use this tool to improve language learning among students. Considering that the type of text has a great impact on the quality of machine translation and very good scientific texts and very bad literary texts are machine translated, this point should also be considered in generalizing the results of this research. All three texts translated by Google were able to match the human translated text in terms of comprehension, but the number of unknown sentences in this text was more than the other two texts, which were expected to have a negative effect on students' comprehension, which was not observed. The issue of gender can also be examined to see if there is a relationship between gender and the type of reaction to machine translation or not.