This repository contains code for a classifier built using the RoBERTa model from Hugging Face's transformers library. The classifier is used to predict the outcomes of legal cases.
The data for this project consists of legal cases, with each case having the following attributes:
- ID: A unique identifier for each case.
- first_party: The first party involved in the legal case.
- second_party: The second party involved in the legal case.
- facts: A summary of the facts of the legal case.
- first_party_winner: A binary attribute indicating whether the first party won the case (1 indicates a win, 0 indicates a loss). This is the target attribute that the model will predict.
Here's an example of the data:
ID | first_party | second_party | facts | first_party_winner |
---|---|---|---|---|
TRAIN_0000 | Phil A. St. Amant | Herman A. Thompson | "On June 27, 1962, Phil St. Amant... | 1 |