improve_vectorspace: Improve a vectorspace by removing common elements.

In bmschmidt/wordVectors: Tools for creating and analyzing vector-space models of texts

Improve a vectorspace by removing common elements.

Description

See reference for a full description. Supposedly, these operations will improve performance on analogy tasks.

Usage

improve_vectorspace(vectorspace, D = round(ncol(vectorspace)/100))

Arguments

vectorspace	A VectorSpacemodel to be improved.
D	The number of principal components to eliminate.

Value

A VectorSpaceModel object, transformed from the original.

References

Jiaqi Mu, Suma Bhat, Pramod Viswanath. All-but-the-Top: Simple and Effective Postprocessing for Word Representations. https://arxiv.org/abs/1702.01417.

Examples

closest_to(demo_vectors,"great")
# stopwords like "and" and "very" are no longer top ten.
# I don't know if this is really better, though.

closest_to(improve_vectorspace(demo_vectors),"great")

bmschmidt/wordVectors documentation built on June 2, 2022, 3:53 p.m.