NEXAJudicia: Structured Dataset for Explainable Legal Judgment Prediction in the Indian Judicial System

This dataset contains structured records of Indian judicial cases used for research on explainable artificial intelligence in legal judgment prediction. The dataset was created to support the development and evaluation of the NEXAJudicia framework, a transformer-based explainable AI system designed to analyze legal documents and predict judicial outcomes while enabling interpretability and bias auditing.

The dataset consists of 4,001 judicial case records collected from publicly accessible Indian court judgment databases covering the period from 2016 to 2025. Each record includes structured legal information extracted from court documents, including case identifiers, narrative descriptions, legal metadata, and final verdict labels.

Key dataset fields include:
- CASE_ID
- CASE_TITLE
- FACTS_SUMMARY
- ISSUES_RAISED
- FULL_TEXT
- ACTS_SUMMARY
- IPC_SECTIONS_USED
- FINAL_VERDICT

The dataset is intended for research in legal natural language processing, explainable AI, legal judgment prediction, and fairness-aware machine learning. It supports experiments involving text classification, interpretability analysis, and bias evaluation in legal decision-support systems.

This dataset accompanies the research article:

"A Novel Transformer-XAI Framework with Bias Attribution and A-RAG for Indian Judicial System."

The dataset is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0), allowing reuse, distribution, and modification provided appropriate credit is given to the authors.