/curiam

Corpus data, models, and analysis for "Corpus re Interpretation and Metalanguage in Supreme Court Opinions"

Primary LanguageJupyter Notebook

CuRIAM

Corpus data and analysis for CuRIAM.

Annotation example

CuRIAM stands for Corpus re Interpretation and Metalanguage. For information about the corpus, see the paper:

Corpus re Interpretation and Metalanguage in Supreme Court Opinions (arXiv, to appear at LREC-COLING 2024)

Corpus Data

The full corpus data is available here.

Setup

  1. Create a conda environment.

    $ conda env create -f environment.yml
$ conda activate curiam

  2. Install the curiam package locally.

    $ pip install --upgrade build
$ pip install -e .

  3. If you plan on running the gamma agreement calculations, install pygamma-agreement separately.

    $ sudo apt-get update
$ sudo apt install coinor-libcbc-dev
$ pip install "pygamma-agreement[cbc]"

    For Apple Silicon, Installing cvxopt via conda and then pip install pygamma-agreement without cbc may work, but I haven't tested it. If it does work, it may be slow.

Acknowledgments

TODO

License

TODO