We illustrate and explain problems of n-grams-based machine translation (MT) metrics (e.g. BLEU) when applied to morphologically rich languages such as Czech. A novel metric SemPO...
This paper presents a practical machine translation system based on sentence types for economic news stories. Conventional English-to-Japanese machine translation (MT) systems whi...
The paper describes a particular approach to multiengine machine translation (MEMT), where we make use of voted language models to selectively combine translation outputs from mul...
We propose a backoff model for phrasebased machine translation that translates unseen word forms in foreign-language text by hierarchical morphological abstractions at the word an...
A translation is a conversion from a source language into a target language preserving the meaning. A huge number of techniques and computational approaches have been experimented...