CMU Multilingual NLP - Dependency Parsing

CMU Multilingual NLP - Dependency Parsing

Graham Neubig via YouTube Direct link

Shift Reduce Example

9 of 26

9 of 26

Shift Reduce Example

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

CMU Multilingual NLP - Dependency Parsing

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Two Types of Linguistic Structure
  2. 2 Why Dependencies?
  3. 3 Universal Dependencies Treebank Standard format for parse trees in many languages
  4. 4 Adding Inductive Bias to Neural Models • Bias self attention to follow syntax
  5. 5 Understanding Language Structure Example of extracting morphological agreement rules using dependency relations
  6. 6 Searching over Parsed Corpora Search using 'syntactic regex'
  7. 7 Analysis of Other Linguistic Phenomena • Examining power and agency in film scripts
  8. 8 Arc Standard Shift-Reduce Parsing (Yamada & Matsumoto 2003, Nivre 2003)
  9. 9 Shift Reduce Example
  10. 10 Classification for Shift-reduce
  11. 11 Encoding Stack Configurations w/ RNNS
  12. 12 Transition-based parsing State embeddings
  13. 13 (First Order) Graph-based Dependency Parsing
  14. 14 Graph-based vs. Transition Based
  15. 15 Chu-Liu-Edmonds (Chu and Liu 1965, Edmonds 1967)
  16. 16 Find the Best Incoming
  17. 17 Subtract the Max for Each
  18. 18 Recursively Call Algorithm
  19. 19 Expand Nodes and Delete Edge Deleted from cycle
  20. 20 Sequence Model Feature Extractors (Kipperwasser and Goldberg 2016)
  21. 21 BiAffine Classifier (Dozat and Manning 2017)
  22. 22 Difficulty In Multilingual Dependency Parsing
  23. 23 Example Improvement 1: Order-insensitive Encoders . Standard cross-lingual transfer can fail with large word order differences between source and target Change model structure to be order-insensitive…
  24. 24 Generative Model Fine-tuning • Use generative model that can be trained unsupervised, and fine-tune on the target language
  25. 25 Example Improvement 3: Linguistically Informed Constraints • Add constraints based on a priori knowledge of the language structure
  26. 26 Discussion Question

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.