GitHub - qshuang123/biomedical-word-embeddings: Exploring vector representations of words in biomedical text

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
.gitignore		.gitignore
PubMed-and-PMC-ri.sh		PubMed-and-PMC-ri.sh
PubMed-and-PMC-w2v.sh		PubMed-and-PMC-w2v.sh
README.txt		README.txt
accuracy.json		accuracy.json
accuracy_PubMed-and-PMC-ri.csv		accuracy_PubMed-and-PMC-ri.csv
accuracy_PubMed-and-PMC-w2v.csv		accuracy_PubMed-and-PMC-w2v.csv
analogies.txt		analogies.txt
analogy.sql		analogy.sql
analyze_accuracy.py		analyze_accuracy.py
calc_accuracy.py		calc_accuracy.py
calc_tsne.py		calc_tsne.py
convert_wvlib.py		convert_wvlib.py
create_analogies.py		create_analogies.py
create_umls_db.sh		create_umls_db.sh
data_sample.csv		data_sample.csv
download_embeddings.sh		download_embeddings.sh
get_vocab.py		get_vocab.py
install_umls.sh		install_umls.sh
install_wvlib.sh		install_wvlib.sh
mrrel_vocab_excise.csv		mrrel_vocab_excise.csv
mrsty_vocab.csv		mrsty_vocab.csv
pairs.csv		pairs.csv
reduce_vocab.py		reduce_vocab.py
requirements.txt		requirements.txt
sample.py		sample.py
tsne_PubMed-and-PMC-ri.csv		tsne_PubMed-and-PMC-ri.csv
tsne_PubMed-and-PMC-w2v.csv		tsne_PubMed-and-PMC-w2v.csv
umls.sql		umls.sql
vocab.csv		vocab.csv
word2vec_util.py		word2vec_util.py

Repository files navigation

biomedical-word-embeddings
==========================

bash download_embeddings.sh
bash PubMed-and-PMC-w2v.sh

UMLS
====

install_umls.sh: downloads and installs the UMLS
create_umls_db.sh: creates a SQLite database containing a subset of the UMLS

Word embeddings
===============

download_embeddings.sh: downloads word embeddings trained on PubMed and PMC text (http://bio.nlplab.org/)

Analogies
=========

analogy.sql: exports a spreadsheet of term pairs with the relationship procedure_has_excised_anatomy
mrrel_vocab_excise.csv: a spreadsheet of term pairs with the relationship procedure_has_excised_anatomy
pairs.csv: cleaned up version of term pairs with the relationship procedure_has_excised_anatomy
create_analogies.py: produces analogies based on pairs.csv
analogies.txt: analogies based on pairs.csv

Install
=======

pip install -r requirements.txt
bash install_wvlib.sh