README.md

metadata

language: en
pretty_name: IndicSupremeCourtRulings
tags:
  - law
  - legal
  - india
  - supreme-court
license: other

IndicSupremeCourtRulings

Overview

IndicSupremeCourtRulings is a structured metadata dataset of Supreme Court of India judgments, organized by year and month-range. Each row corresponds to a judgment metadata record and includes a link to the original judgment PDF.

Source

The metadata in this dataset was collected from the official Supreme Court of India website and associated endpoints. This dataset does not redistribute the PDF files unless they are already present locally in judgment_pdf_path. The canonical source for each judgment is the judgment_url.

Important: Legal/Policy Compliance (India)

By using this dataset, you agree to comply with all applicable Indian laws and regulations governing access, use, reproduction, and redistribution of judicial materials and any associated documents.
You are solely responsible for ensuring that your use of the dataset (including any downloading, copying, processing, or redistribution of judgment PDFs) is lawful in your jurisdiction and under Indian law.

Dataset Structure

• Unit of observation: one metadata record per judgment
• Key fields: case_number, diary_number, petitioner_respondent, bench, judgment_url, date_from, date_to

Fields

See dataset_info below for the full schema.

---

dataset_info: features: - name: serial_number dtype: string - name: diary_number dtype: string - name: case_number dtype: string - name: petitioner_respondent dtype: string - name: advocates dtype: string - name: bench list: string - name: judgment_by dtype: string - name: judgment_url dtype: string - name: judgment_pdf_path dtype: string - name: date_from dtype: string - name: date_to dtype: string - name: year dtype: int32 - name: month_range dtype: string - name: source_file dtype: string splits: - name: train num_bytes: 16725850 num_examples: 38603 download_size: 3159873 dataset_size: 16725850 configs: - config_name: default data_files: - split: train path: data/train-*