It's surprising, then, how quickly the researchers were able to achieve this milestone – especially given that machine translation is a problem people have been trying to solve for decades.
When faced with labels in different languages, most approaches resort to general-purpose machine translation services to reduce the problem to monolingual English-only matching.
Google uses related approaches to tackle a number of more practical problems, such as machine translation and voice recognition.
But much of the original impetus for this work disappeared, for a time at least, in part because of the realization that the theoretical problems involved in machine translation are much more difficult than they were at first thought to be and in part as a consequence of a loss of interest among linguists in the development of discovery procedures.
That is, one that isn't a binary choice between machine translation, such as Google Translate, or simply throwing humans who speak more than one language at the problem.
Developments in the treatment of translating natural language as a machine learning problem (known as statistical machine translation - SMT) are leading to continual improvements in this field although the linguistic accuracy varies based on the way each machine 'learns' (Lopez, 2008).
It begins with a description of the many challenging problems faced by designers of machine translation systems and explains why there appears to be slim hope of attaining human-level performance in the foreseeable future.
One of the many problems in the field of machine translation is that expressions (multi-word terms) convey ideas that transcend the meanings of the individual words in the expression.
The research that addresses these problems represents the first real task-driven focus since machine translation research in the 1960s.
To minimize these problems, in this paper, we propose an architecture for machine translation to Brazilian sign language (LIBRAS) and its integration, implementation and evaluation for digital TV systems, a real-time and open-domain scenario.
