Legal Statute Identification: A Case Study using State-of-the-Art Datasets and Methods

Overview

This repo contains the material for our paper.

1. Codes for all 8 discriminative models, 2 llm prompting models, as well as for anonymization, building LADAN commmunities and calculating confusion are present in "codes" folder.
2. All dataset related files can be found inside ILSI and ECHR folders.
   • Some materials which could not be provided in GitHub are present here.
3. All metrics, predictions and auxillary outputs are present inside "models" folder, which contains ILSI and ECHR subfolders.
   • The trained models could not be provided in GitHub and can be found here.