Статистический машинный перевод и нечеткая логика?

Question

[w495]

Связан ли как-то статистический машинный перевод с нечеткой логикой? Я чувствую что да, но никак не могу понять, как именно.

Очень извиняюсь за вопрос, но мне не дает покоя, а рунете не так много мест, где можно получить на него ответ.

Answer 1

[rasa]

Возможно, вам поможет книга «Routledge encyclopedia of translation studies»

В частности, в ней есть такие строки:

<sup>One advanced feature of a translation memory system is the so-called fuzzy match facility. This is of interest to both professional translators and linguists. In addition to exact matches, systems which incorporate this feature can find in their translation memory segments which differ from each other in certain respects but which are regarded as similar according to specific algorithms. The algorithms are based on the principles of ‘fuzzy logic’ and make use of syntactic parsing mechanisms to some extent.

A similar principle of ‘fuzzy match’ or ‘controlled similarity’ has been applied in a rather different translation environment. A system known as Computer-Aided Dialect Adaptation (CADA) has been used by Bible translators to create a translation on the basis of another translation in a similar or related language, for example to translate the Scriptures into several dialects of South American or African languages (Bean 1993; Stanford and Watters 1993).</sup>

в html

Comments

[w495]

Спасибо