Svitlana Hrushko
Condidate of Philology, Associate Professor of Translation and Theoretical
and applied linguistics of the State Institution «South Ukrainian National Pedagogical University named after K. D. Ushinsky»
Odesa, Ukraine
e-mail: svetagrushko64@gmail.com
ORCID ID https://orcid.org/0000–0001–9940–7780
DOI: https://doi.org/10.24195/2616-5317-2020-30-4
Key words: machine translation, algorithm, linguistic equivalents, models of translation.
The purpose of the article is to study problems of translation equivalence in machine translation, which is based on a sequence of invariable actions (algorithms) with a text to identify linguistic equivalents in a pair of languages at a given direction of translation by means of a computer, in respect of the pragmatic aspect.
Translation equivalence is understood as a specific type of equivalence, which is fundamentally different from other types, since it does not correlate with the phenomena that have a special place in the structure of a language, but the phenomena that currently exist in a language correlation or are equivalent to the text content. The translation is formalized, but allows getting an idea of the text content at the introductory level, since it is not an accurate, adequate translation, but performs the function of rendering basic information. Machine translation is not able to render nuances of an original text, not only at the lexical level. When translating, it is necessary to take peculiarities of syntax and semantics into account. Adequate computer translation is almost impossible in this case. This fact is recognized by all scholars who study possibilities of this type of translation only when rendering main content of a document without taking language nuances and features into account.
Machine translation can be carried out on a basis of the translation equivalence (objective and dynamic) model. The model in terms of linguistic technology provides an optimal solution of problems of independent linguistic description and algorithm.
The system of translation equivalence, which can be implemented within the model of translation equivalence, allows providing sufficient quality of machine translation at the pre-editing stage. When creating a machine translation program, in addition to solving linguistic problems, a program of their implementation is also necessary, since a translation program is a tool for studying and finding information in a foreign language, and the prospects of a machine translation are related to the further development of translation theory and practice in general.
REFERENCES
- Valipur, A. (2010). Neskolko slov ob ispolzovanii formalizatsii i funktsionalizma pri obrabotke teksta [A Few Words on Formalization and Functionalism in Text Processing]. www.ensani.ir. Retrived from https: // www.ensani.ir/storage/Files/20101208140318–308.pdf [in Russian].
- Kotov, R. G., Novikov, A. I., & Skokan, Yu. P. (1987). Prikladnaia lingvistika i informatsionnaia tekhnologiia [Applied Linguistics and Information Technology]. Moscow: Nauka [in Russian].
- Marchuk, Yu. N. (1983). Problemy mashinnogo perevoda [Machine Translation Issues]. Moscow: Nauka [in Russian].
- Marchuk, Yu. N. (2007). Kompiuternaia lingvistika [Computational Linguistics]. Moscow: AST: Vostok-Zapad [in Russian].
- Marchuk, Yu. N. (1979). Sintaktiko-semanticheskii analiz v sisteme mashinnogo perevoda AMPAR [Syntactic-semantic Analysis in the AMTER Machine Translation System]. Proceedings from Mezhdunarodnyi seminar po mashinnomu perevodu — International Workshop on Machine Translation. (pp. 8–9). Moscow: VTsP [in Russian].