Machine translation

1. What is Machine Translation (MT)?

• a. Word-based translation (historical beginning)

• b. Phrase-based translation (current state implementation in the business)

• c. Hierarchical-phrase-based machine translation (research state-of-the-art)

2. MT in business settings == post-editing (of phrase-based MT) or raw translation provision

• a. MOSES (quick MOSES DIY – I might skip this part and provide the necessary links, where more can be read)

• b. A different use case – Unbabel.

• c. Learn from human-translators’ feedback

3. Evaluation (continuous v.s isolated)

• a. We are the first ones to speak about continuous MT evaluation to the science community

4. Challenged and Unsolved Issued

• a. Translation of low-resource language pairs

• b. Out-of-domain text

• c. Morphological issues

• d. Terminology inconsistency

• e. Formatting issues

• f. Word order



The presentation will be given by Ekaterina Stambolieva.

For first time comers here is the access plan: 

The event is free of charge.


Join or login to comment.

  • Ekaterina S.

    Hi, sorry I took long to get back to you with what I have promised as a follow-up.
    Here it goes:

    MOSES DYI -> http://www.statmt.org/moses/?n=Moses.Baseline

    Speaker's notes about MT in general -> http://moreintelligent.wordpress.com/2014/04/01/smt/

    Speaker's notes about IBM1 implementation details (I still have not found time to finish this, what is left is to go through the calculation itself for the first iteration + the language model) -> http://moreintelligent.wordpress.com/2014/04/10/learn-word-to-word-translations-ibm1/ Do not hesitate to contacting me back if you want to!
    Have a great start of the week!

    Cheers,
    Ekaterina

    April 13, 2014

  • Thomas D.

    The talk and presentation was very interesting well delivered:)
    Looking forward to learn more!

    Especially the explanation of the IBM Model 1 algorithm was very enlightening!

    April 2, 2014

  • Hugues M.

    고맙습니다 and very interesting

    April 1, 2014

  • Ekaterina S.

    Change of presentation outline:
    1. What is MT?
    • a. Word-based translation (historical beginning)

    • b. Phrase-based translation (current state implementation in the business)

    • c. Hierarchical-phrase-based machine translation (research state-of-the-art)

    2. Lexical Translation and the Expectation Maximization algorithm

    3. Language Modelling (quickly saying what is this all about)

    4. (If there is time) Evaluation (continuous v.s isolated)


    • a. We are the first ones to speak about continuous MT evaluation to the science community

    4. (If there is time) Challenged and Unsolved Issued


    • a. Translation of low-resource language pairs

    • b. Out-of-domain text

    • c. Morphological issues

    • d. Terminology inconsistency

    • e. Formatting issues

    • f. Word order

    See you tomorrow.

    March 31, 2014

  • Peadar C.

    I'd love to attend. But unfortunately I've now moved to Brussels... So I doubt I'll make it to many meetups in Luxembourg.

    March 31, 2014

  • Dzidorius M.

    Attention! The date has been moved from 2nd to 1st of April.

    1 · January 27, 2014

16 went

Our Sponsors

People in this
Meetup are also in:

Sometimes the best Meetup Group is the one you start

Get started Learn more
Katie

I'm surprised by the level of growth I've seen since becoming an organizer, it's given me more confidence in my abilities.

Katie, started NYC ICO

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy