Clone all the submodules: git submodule update --init --recursive
required python2
- create a venv
- install required things by using
install.sh
- download the models for open-sesame and semafor: run
get_models.sh
- get the corpora: run
get_corpora.sh
- extract the sentences from the corpora: run
text_extract.py
- annotate with the prebuilt systems: run
annotate.sh
- group the annotations on the same sentences: run
aligner.py
- extract the TSV files for easier comparison: run
to_tabular.py
automate the framenet source files download?? need:
- FrameNet:
data/fndata-1.7/luIndex.xml
- preprocessed conll:
data/neural/fn1.7/fn1.7.fulltext.train.syntaxnet.conll
- embeddings:
glove.6B.100d.txt