Informace o projektu

Anotace je dostupná pouze v anglickém jazyce.

This project will examine the current methods for finding translation equivalents without using parallel texts. The goal is to evaluate the current methods, choose and/or adjust the most suitable one and determine the efficient parameters of the word-to-word alignment datasets used for supervision. This project will produce: datasets for the evaluation, evaluation of the current methods, analysis of the word-to-word alignment datasets, and method for efficient finding translation equivalents.