Building and Developing a Parallel Corpus between Turkish and Arabic for The Purposes of Teaching Arabic to Turkish Native Speakers

Abstract

The idea of this study is to try to connect between the ability to acquire the second language and the society's nature in which the language is used. The researcher seeks to apply his idea on the model of teaching Arabic to Turkish native speakers, in view of their need to use Arabic in their Turkish society. Because teaching Arabic language requires understanding the reality of its usage, whether that usage in its Arab society or in another society, this study seeks to provide a methodological vision of how to build a bilingual parallel corpus between Turkish and Arabic, then process this corpus to extract the the most suitable knowledge of Arabic vocabulary and structures adapted to Turkish learner society. The study will be guided by the Common European Frame- work of Reference for Languages criteria. This research is based on a descriptive approach studying language usage in its two dimensions: quantitatively and qualitatively. It is the way combining both of corpus linguistics and educational linguistics. Among the aims of the study is to conciliate between educational books and the educational environment where the student finds himself forced to learn strange and rare vocabulary, structure and styles that he doesn't need, which affects his ability to practice and master language. Among the goals achieved by the research is developing a parallel corpus between Turkish and Arabic language, processing its material in order to teach Arabic to Turkish native speakers, providing a methodological vision of how to use this corpus in teaching Arabic as a second language and providing an easy way that helps to benefit from educational resources in saving time and effort spent in teaching and learning. an interactive model for the learner's environment, then generating the target language through translation. The researcher exempli-fied this by the model of Turkish society when it seeks to learn Arabic in order to use it in this society. The study presented the problems of constructing a parallel corpus for the purposes of teaching Arabic. These problems were represented in scarcity of digital Arabic content, disparity of the linguistic system between Arabic and Turkish and the recent use of linguistic corpora in the fields of educational linguistic research. At the application level, it was possible to extract the Arabic texts, and the researcher subjected the Arabic texts to statistical processing in order to prepare a list of unique words. The number of these words reached (295,654) words from among the Arabic words, with a percentage of 3%. At the application level also, it was possible to subject Arabic texts to computational processing. This processing resulted in multiple outputs, including: the most revolving particles, the most revolving words, the most revolving col-locations, models of dialogic texts, models of descriptive texts and models of texts and annotations. The researcher presented a vision of how to use these outcomes in teaching Arabic to Turkish speakers at levels A, B and C. The study presented a vision for future work that aims to solve the deficiencies through measures, including: developing the corpus to include multiple fields, testing the corpus efficiency to ensure that its material represents society and investing outputs in producing useful educational resources.