The decision ambiguity problems for the Kazakh language

Abstract

This paper describes the problems of word ambiguity for the Kazakh language, which is one of the languages ​​of the Turkic class. The process of solving the problem of ambiguity for the Kazakh language in the machine translation system for the Kazakh - English and Kazakh - Russian language pairs (and vice versa) is described on the basis of the proposed technology. The proposed technology includes a constraint grammar model and a maximal entropy model for more effectively solving the problem of lexical selection for the Kazakh language. Results are presented by comparing two technologies separately and together.