Code for the paper Unsupervised Transfer of Semantic Role Models from Verbal to Nominal Domain by Yanpeng Zhao and Ivan Titov.

Our model transfers the argument roles of verbal predicates (acquired) to the arguments of their nominalization (acquisition). acquired and acquisition share the lemma. Note that we do not rely on any argument alignment.

To be added...

The SRL development environment relies on a specific commit of AllenNLP and needs a patch to fix a compatibility issue of that commit.

git clone --branch beta https://github.com/zhaoyanpeng/srl.git
cd srl
virtualenv -p python3.7 ./pyenv/oops
source ./pyenv/oops/bin/activate
pip install --upgrade pip
pip install --upgrade setuptools
pip install -r requirements.txt

ALERT: this is an outdated repo. Before installing AllenNLP, maybe first resolve the mentioned issue (see Resolve the AllenNLP Issue section); unfortunately, I forget what exactly the issue is.

git clone https://github.com/allenai/allennlp.git
cd allennlp
INSTALL_TEST_REQUIREMENTS=true scripts/install_requirements.sh
pip install --editable .
cd ..

git checkout 11d8327 -b patch
git checkout 385e66e -b srl_base
git checkout patch -- allennlp/semparse/domain_languages/domain_language.py
git branch -d patch

./executor/srl_toy.sh

@misc{zhao2020unsupervised,
      title={Unsupervised Transfer of Semantic Role Models from Verbal to Nominal Domain}, 
      author={Yanpeng Zhao and Ivan Titov},
      year={2020},
      eprint={2005.00278},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://doi.org/10.48550/arXiv.2005.00278}
}

MIT