COMPARISON OF THE PARSING TREES STRUCTURE IN RELATION TO ENGLISH-JAPANESE TRANSLATION

This article focuses on the parsing trees in natural language processing specifically for the English and Japanese languages for the purpose of translation between those two languages. Throughout this article, the parsing trees are built for the original phrase from the speech “I have a dream” by Martin Luther King Jr and its Japanese translation. Then these parsing trees are compared and the differences between the structures are highlighted for each language. Finally, the sequence of procedures is proposed for both languages with the emphasis on the highlighted differences between those two.The purpose of this study is to determine the peculiarities of parsing trees construction for the English and Japanese languages and to suggest a basic set of principles for the conversion of parsing trees between the two languages.The topicality of this study lies in the ever-increasing need for quality machine translation due to the drastically rising amount of data processed daily and the global tendency for digitalization additionally spurred by the ongoing pandemic.During this study, the following methods were implemented - scientific modeling, comparative and analytical. The method of scientific modeling was applied to the text data in both languages to obtain the parsing trees for comparison. During the next step, the comparative method was employed to compare the resulting structures in both languages and analyze their differences. Lastly, the analytical method was used to evaluate the resulting characteristics and to propose the set of structure transformation methods.The results of this study consist of the set of structure transformation methods proposed for the conversion of the sentence structure when translating both from the English language to Japanese and vise versa. This algorithm is heavily influenced by the difference in the word order of languages and therefore resulting parsing trees happen to have right-branching main structure and left-branching for the modifiers in the English language and prominent left-branching movement in the Japanese language. Both this movement controversy and the lack of typical for alphabetic languages clear word boundaries, which have to be deduced in the Japanese language based on the inflections, lead to operation difficulties while translating between the two languages. These difficulties were expectantly reduced in the course of this study.