NEURAL MACHINE TRANSFER: OPENNMT USE FOR TRAINING MODEL TRANSLATOR

Authors

        Soliev P.A.Senior teacher, Chair of Programming and Information Technologies  Department, Polytechnic Institute of Tajik Technical University.

       Kosimov A.A. Senior teacher, Chair of Programming and Information Technologies Department, Polytechnic Institute of Tajik Technical University.

Annotation

      The article provides an explanation of the processing module for neural machine translation using OpenNMT. The stages of learning information and the re-preparation of the system for translation are shown. Also mentioned is the processing module for a Tajik machine translator. The results of experiments on the integration of a neural network using OpenNMT into a translator are discussed. The general principle of operation of machine translation programs using OpenNMT and problems related to the adequacy of the translated text when using an electronic translator are described. It is proved that by connecting a neural network to a translator, it can be achieve a more accurate equivalent of the translated word, sentence, and, as a final result, better translation of the entire text of different genres.

Key words

 machine translation, neural network, training, electronic translation, model, OpenNMT.

Language

english

Type

technical

Year

2019

Page

18

References

  1. Soliev P.A., Kosimov A.A. – Neural networks and its use in machine translation // Bulletin of Polytechnic Institute of Tajik Technical University «Scientific-technical journal», Khujand, 2018, № 3 (8), P. 11 – 16.
  2.  OpenNMT: Open-Source Toolkit for Neural Machine Translation. Materials of scientific conferences. Guillaume Kleiny, Yoon Kim, Yuntian Deng, Jean Senellarty, Alexander M. Rush. Harvard University, SYSTRAN.
  3.  An OpenNMT Model to Arabic Broken Plurals. Elsayed Issa. University of Arizona, 845 N Park Ave, Tucson, AZ 85719, USA.
  4.  Improving a Multi-Source Neural Machine Translation Model with Corpus Extension for Low-Resource Languages. Gyu-Hyeon Choi, Jong-Hun Shin, Young-Kil Kim. Korea University of Science and Technology (UST).
  5.  Soliev O.M., Khudoyberdiev Kh.A., Soliev P.A.Web-Application Tajik_English_Translator // Certificate of State Registration of an Information Resource, Republic of Tajikistan, 31.01.2018, № 4201800391.

Date publication

09/21/2023