Computer scienceData scienceNLPLanguage representationEmbeddings

Word2Vec

Theory

Which Methods?

Report a typo

Suppose we want to train a Word2Vec model on movie reviews using NLTK. We specify the following parameters:

w2v_model = Word2Vec(min_count=10,
                     window=10,
                     sg=1,
                     size=300,
                     sample=6e-5)

Once we've done it, we need to construct the dictionary from the corpus and then learn the embeddings:

# constructing a dictionary
w2v_model.___________(movie_reviews.sents())

# learning the embeddings
w2v_model.___________(movie_reviews.sents(), total_examples=w2v_model.corpus_count, epochs=15)

Which methods should we use for this purpose? Write their names in the text field below, in the same order as we would use them to train a model. Separate the names with a single space.

Enter a short text
___

Create a free account to access the full topic