Study of Sequence Labeling Architectures in NLP

Experimental study of State-of-the-art deep learning models for different Sequence Labeling tasks in NLP

• Performed comparitive study of different modeling choices

  1. CRF vs softmax layer
  2. LSTM vs CNN character embeddings
  3. Effect of language models
  4. Effect of size of vocabulary

Github Repository slides