How do Computers Learn a New Language? -- An Introduction to Statistical Machine Translation

This META-NET video takes a look at how Statistical Machine Translation works in practice. The basic concepts and principles are explained in a high-level way as well as some historical background of the ideas behind machine translation and a look at what sort of data SMT systems use and how.

[@easytranslation5242]

Awesome explanation....I agree that machine can do a lot, but not everything...there is always limitations.

[@yankumar5280]

thanks for sharing META-NET

[@lslsls-mu5rb]

Machine translation is all good and well. This is however NOT a replacement of the translation process. In a MT process, it is the Post-Editing (PE) that replaces the translation phase. Language service users are still advised to enforce a reviewing process like you'd normally do with conventional translations.

[@mariwanmt4639]

My major is Machine Translation. Is there anyone who has the same field?

[@oliverli9630]

4:59 that's how translation workforce keep their job. they have more quality translation data than all the publicly available pieces combined, whose models will output far better translations than google and the other giants. E.g., for "Go get it / keep up the good work / keep at it," you'll only see "Come on" spit out from a chinese-english translation LOL

[@LmH89]

Computers 'learn' in this sense by collecting and using patterns from texts fed to them.

[@mebeasensei]

I know it is an introduction but am sorry but I think this is just rehashed common knowledge plus one huge error, that grammar checkers can sort out the problems, before adding the disclaimer at that end that, they can't after all. I want to see an example of how grammar checkers are actually used in conjunction with MT in practice before I watch another one of your videos.

[@inserthere6387]

espcially for lawyers lol international law here I come lol