By calculating a word2vec model on the corpus we can calculate OTHER words that are used similarly to the ones that we are interested in and thereby find more relevant terms.
This could take too long to calculate for a single notebook session -- if so, we need to pre-calculate all the terms and store mappings down to N (5?) depth and go through that list manually.
By calculating a word2vec model on the corpus we can calculate OTHER words that are used similarly to the ones that we are interested in and thereby find more relevant terms.
This could take too long to calculate for a single notebook session -- if so, we need to pre-calculate all the terms and store mappings down to N (5?) depth and go through that list manually.