000 02200nam a22002537a 4500
003 IIITD
005 20240216133422.0
008 240215b xxu||||| |||| 00| 0 eng d
020 _a9781108497329
040 _aIIITD
082 0 0 _a418.020
_bKOE-N
100 1 _aKoehn, Philipp
245 1 0 _aNeural machine translation
_cby Philipp Koehn
260 _aCambridge :
_bCambridge University Press,
_c©2020
300 _axiv, 393 p. :
_bill. ;
_c26 cm.
504 _aIncludes bibliographical references and index.
505 _t1. The Translation Problem
_t2. Uses of Machine Translation
_t3. History
_t4. Evaluation
_t5. Neural Networks
_t6. Computation Graphs
_t7. Neural Language Models
_t8. Neural Translation Models
_t9. Decoding
_t10. Machine Learning Tricks
_t 11. Alternate Architectures
_t12. Revisiting Words
_t13. Adaptation
_t14. Beyond Parallel Corpora
_t15. Linguistic Structure
_t16. Current Challenges
_t17. Analysis and Visualization
520 _a"A decade after the publication of my textbook Statistical Machine Translation, translation technology has changed drastically. As in other areas of artificial intelligence, deep neural networks have become the dominant paradigm, bringing with them impressive improvements of translation quality but also new challenges. What you are holding in your hands started out four years ago as a chapter of an envisioned second edition of that textbook, but the new way of doing things expanded so dramatically, and so little of previous methods is still relevant, that the text has grown into a book of its own right. Besides the chapter on evaluation, there is very little overlap between these two books. For the novice reader interested in machine translation, this is good news. We all started anew a few years ago, so you are not far behind in learning about the state of the art in the field"--
650 0 _aMachine translation.
650 0 _aNeural networks (Computer science)
650 0 _a Subject: Translating and interpreting -- Data processing.
776 0 8 _iOnline version:
_aKoehn, Philipp, 1971-
_tNeural machine translation
_bFirst ed.
_dNew York : Cambridge University Press, 2020.
_z9781108608480
942 _2ddc
_cBK
999 _c172241
_d172241