The Role of Word Embeddings

Publication at https://doi.org/10.5281/zenodo.11348467

Local Installation

To experiment with word embeddings locally, the dataset Gensim could be used with Python 3.

Python 3 can be installed from https://www.python.org/

There is currently a bug in the Gensim module. To fix it, an older version of scipy could be used. See https://stackoverflow.com/q/78279136

pip install scipy==1.12
pip install gensim

Run Examples Locally

python3

import gensim.downloader as api
dataset = api.load('glove-wiki-gigaword-300')

dataset.most_similar_cosmul(positive=["warm", "winter"], negative=["summer"])

dataset.most_similar_cosmul(positive=["boat", "road"], negative=["sea"])