Wiki Lecture 3 – GloVe: Global Vectors for Word Representation


#1

Lecture video

Lecture Notes

Readings:
What is GloVe? Part 1
GloVe: Global Vectors for Word Representation
Improving Word Representations via Global Context and Multiple Word Prototypes
Improving Distributional Similarity with Lessons Learned fromWord Embeddings
Evaluation methods for unsupervised word embeddings

Paper spotlight
Linear Algebraic Structure of Word Senses, with Applications to Polysemy

Assignment:

  • Continue working on assignment 1.
  • Implement Glove in Python.
  • What are the disadvantages of word2vec(skip-gram and CBOW) compared to GloVe? Hit ‘Reply’ and discuss.
  • GloVe, skip gram and Bag of Words are distributed representations. What are distributed representation? How are they more efficient than one-hot encoding? Hit ‘Reply’ and discuss.