cleanNLP (version 1.10.0)

get_vector: Access word embedding vector from an annotation object

Description

Word embeddings map each lemma or token into a high-dimensional vector space. The implementation here uses a 300-dimensional space. Only available with the spaCy parser.

Usage

get_vector(annotation)

Arguments

annotation

an annotation object

Value

Returns a matrix containing one row for every triple found in the corpus, or NULL if not embeddings are present

References

Pennington, Jeffrey, Richard Socher, and Christopher D. Manning. "Glove: Global Vectors for Word Representation." EMNLP. Vol. 14. 2014.