Problems with Cosine as a Measure of Embedding Similarity for High Frequency Words.

Kaitlyn Zhou Kawin Ethayarajh Dallas Card Dan Jurafsky

Published in: CoRR (2022)

Keyphrases

high frequency
low frequency
similarity measure
distance measure
high resolution
euclidean distance
wavelet transform
visual quality
high frequencies
wavelet coefficients
vector space
low pass
machine learning
high quality
high frequency components