This repository is the official implementation of Semantically-informed Hierarchical Event Modeling (Published in *SEM 2023).
git clone https://github.com/dipta007/SHEM
cd SHEM
git checkout ind_frame
conda create -n shem python=3.7
conda activate shem
conda install pip
conda install pytorch torchvision torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia
pip install torchtext==0.2.3
pip install -U scikit-learn
conda install -c conda-forge pytablewriter
conda install -c anaconda pandas
pip install gdown
conda activate shem
pip install gdown
mkdir saved_models
mkdir saved_configs
gdown https://drive.google.com/drive/u/1/folders/1YUfHGdyVXONtvbJQCbBN0OA0ibXP82bw -O ./data --folder
There are in total 10 frame relations. To change the frame relation, change the value of the variable frame_relation in ./framenet_relations.py to one of the following values:
frame_relation = 'Inheritance'
frame_relation = 'Using'
frame_relation = 'Precedes'
frame_relation = 'Metaphor'
frame_relation = 'See_also'
frame_relation = 'Causative_of'
frame_relation = 'Inchoative_of'
frame_relation = 'Perspective_on'
frame_relation = 'Subframe'
frame_relation = 'ReFraming_Mapping'
./train.sh $obsv_prob $exp_num $seed
{'valid','test'}
./test_ppx.sh $obsv_prob $exp_num $seed $data_mode
./wiki_inv_narr.sh $obsv_prob $exp_num $seed $data_mode
Some parts of the code were inspired by SSDVAE implementations.