Successfully reported this slideshow.

20 Slides

180 views

Published on

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

20 Slides

  1. 1. NLP for minority languages <ul><li>Chris Brew </li></ul><ul><li>CSE and Linguistics </li></ul>
  2. 2. Anna Feldman <ul><li>faculty at Montclair State University </li></ul><ul><li>Russian native speaker. Slavic language expertise </li></ul><ul><li>Learnt to program while at OSU </li></ul><ul><li>Double major BA in English and East Asian Studies, HUJ </li></ul>
  3. 3. Russian POS tagging Jan Hajič Eva Hajič ova
  4. 4. Resource light tagging <ul><li>All human languages are related in some way. </li></ul><ul><li>Some of them closely related. </li></ul><ul><li>We used that fact </li></ul>
  5. 6. Markov Model Det Adj Noun A red book
  6. 7. Markov Model Adj-Nom Nn-Nom книга красная
  7. 8. Markov Model Adj-Nom Nn-Nom kniha Červená Shared
  8. 9. Markov Model Adj-Nom Nn-Nom kniha Červená Shared
  9. 10. Markov Model Need something to get Russian words Adj-Nom Nn-Nom kniha Červená Shared
  10. 11. Text
  11. 13. Text
  12. 14. <ul><li>Kirk Baker </li></ul><ul><li>Phonetician/phonologist, worked with Chip Gerfen at UNC </li></ul><ul><li>Now working at Collexis in DC area as computational linguist. </li></ul>
  13. 15. Heuristics for animacy
  14. 16. Tagalog: not yet in Google Translate
  15. 17. Credo <ul><li>Language provides computer science with very interesting problems </li></ul><ul><li>Theoretical ideas like language relatedness can have real technological impact </li></ul><ul><li>Machine learning has a kind of eldritch magic </li></ul><ul><li>Teach me about your beautiful language... </li></ul>

×