Machine Translation

Introduction

1. What is probably the most common use of machine translation today?
   Information access: translating web instructions, recipes, news articles, Wikipedia or government pages written in another language.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 13, Introduction, p. 276
2. What is the digital divide in information access, and how can MT help?
   Much more information is available in English and other languages of wealthy countries (more search results, bigger Wikipedias). Good MT can bring that information to speakers of lower-resourced languages.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 13, Introduction, p. 276
3. What are computer-aided translation (CAT), post-editing and localization?
   CAT: MT produces a draft that a human translator fixes in a post-editing phase. Localization: adapting content or a product to a particular language community, commonly using CAT.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 13, Introduction, p. 276
4. Name newer, in-the-moment uses of MT.
   Incremental translation of speech on the fly before the sentence is complete (as in simultaneous interpretation), and image-centric translation, e.g. OCR of menus or street signs from a phone camera.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 13, Introduction, p. 276
5. Why is MT naturally modeled with an encoder-decoder (sequence-to-sequence) architecture?
   The output is a complex function of the entire input: target words don’t correspond to source words in number or order (Japanese puts the verb last and drops the pronoun “he”).

   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 13, Introduction, pp. 276–277