/language-models-are-knowledge-graphs-pytorch

Language models are open knowledge graphs ( non official implementation )

Primary LanguagePythonMIT LicenseMIT

language-models-are-knowledge-graphs-pytorch

Language models are open knowledge graphs ( work in progress )

A non official reimplementation of Language models are open knowledge graphs

The implemtation of Match is in process.py

example bob dylan

Execute MAMA(Match and Map) section

Do note the extracted results is still quite noisy and should then filtered based on relation unique pair frequency

python extract.py examples/bob_dylan.txt bert-large-cased-bob_dynlan.jsonl --language_model bert-large-cased --use_cuda true

Map

  1. Entity linking

The original download link for Stanford Entity linking is removed (nlp.stanford.edu/pubs/crosswikis-data.tar.bz2)[nlp.stanford.edu/pubs/crosswikis-data.tar.bz2]. I will use (REL)[https://github.com/informagi/REL] for entity disambiguation model (supervised instead of the original unsupervied) to achieve the same task.

  1. Relations linking (page 5, 2.2.1)

Lemmatization is done in the previous steps process.py, in this stage we remove inflection, auxiliary verbs, adjectives, adverbs words.

Adjectives extracted from here: https://gist.github.com/hugsy/8910dc78d208e40de42deb29e62df913

Adverbs extracted from here : https://raw.githubusercontent.com/janester/mad_libs/master/List%20of%20Adverbs.txt

Environment setup

This repo is run using virtualenv

virtualenv -p python3 env
source env/bin/activate
pip install -r requirements.txt