Exploring NLP with Coursera - Part 5

Sequence tagging with Probabilistic Models

Basically sequence tagging is used in real-world use cases where for example we need to find named entities in a sentence (e.g. Person, City, Date) or find out parts of speeches (e.g. Verb, Noun, Adj, etc.).

There are various approaches to sequence labeling:

  • Rule-based models (not covered in this course)
  • Separate label classifiers for each taken (classification approachh)
  • Sequence models (HMM, MEMM, CRF) - topic of this lecture
  • Neural Networks

Hidden Markov Models