To the problem of text classification in machine translation

Abstract

Classification of texts liable for automatic
translation is of crucial theoretical and practical
importance for increasing Machine Translation
productivity. The article considers main principles to
take into account when efficient classification is being
developed. Necessity of making classification on
the basis of texts lexis is substantiated. Its also noted
that possible solution to the problem is through detailed
description of texts lexical groups and necessity
to use this data when machine dictionaries are
being designed for machine translation system.