Word embeddings from Pubmed and PMC

[gccome]

Can we consider adding word embeddings for biomedical filed? Various related w2v can be downloaded at http://evexdb.org/pmresources/vec-space-models/

More intro can be found at http://bio.nlplab.org

[alanakbik]

Good point! Yes, should be added!

[bharatr21]

@gccome PRs #519 and #503 already added the FlairEmbeddings('pubmed-forward') ,
FlairEmbeddings('pubmed-backward') and ELMoEmbeddings('pubmed') which I'm guessing is already from http://bio.nlplab.org in this link, but support for PMC, Pubmed+PMC and Wikipedia+Pubmed+PMC could be added, at the cost of making FlairEmbeddings or ELMoEmbeddings of those corpora very large (since the last two are more than 2 GB).
@alanakbik Any thoughts on adding support for the last two?

[gccome]

@Bharat123rox thanks for the info! What I meant is that we can call something like WordEmbedding('pubmed'). And I think FlairEmbeddings('pubmed-forward'), FlairEmbeddings('pubmed-backward'), and ELMoEmbeddings('pubmed') don't explicitly use those three word embeddings from bio.nlplab.org. @alanakbik is it correct?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.