statistical machine translation
[/stəˈtɪstɪkəl məˈʃiːn trænzˈleɪʃən/]
nounpl: statistical machine translations
tradução automática estatística
1. A machine translation approach that uses statistical models trained on large parallel corpora to translate text from one language to another, based on probability distributions rather than explicit linguistic rules.
Statistical machine translation was the dominant approach in the field before the rise of neural machine translation.
A tradução automática estatística foi a abordagem dominante no campo antes do surgimento da tradução automática neural.
2. A computational method that learns translation patterns from bilingual text data and applies probabilistic models to generate translations.
Google Translate utilized statistical machine translation for many years before switching to neural approaches.
O Google Tradutor utilizou tradução automática estatística por muitos anos antes de migrar para abordagens neurais.
This is a highly specialized technical term primarily used in computational linguistics, computer science, and translation technology sectors. It emerged as a major field of research in the late 1990s and early 2000s, particularly at major tech companies and universities in both the United States and Brazil. The term reflects the evolution of machine translation technology and remains important for understanding the history of AI and natural language processing, even as neural machine translation has become more prevalent in recent years.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →