Abstract: This paper presents \texttt{PredEx}, the largest annotated dataset for legal judgment prediction and explanation in the Indian context. This unique corpus enhances the training and evaluation of AI models in legal analysis. Our work innovates by applying instruction tuning to Large Language Models (LLMs), significantly improving their predictive accuracy and explanatory depth for legal judgments. We employed various transformer-based models, tailored for both general and Indian legal contexts. Through a combination of lexical, semantic, and expert assessments, we demonstrate the effectiveness of our approach. Despite challenges like handling extensive documents and reducing hallucinations, our results are promising, indicating a significant leap forward in AI-assisted legal judgment prediction and explanation. This study not only contributes a groundbreaking dataset but also paves the way for future advancements in AI-assisted legal judgment prediction and explanation.
Paper Type: long
Research Area: Resources and Evaluation
Contribution Types: Model analysis & interpretability, Publicly available software and/or pre-trained models, Data resources, Data analysis
Languages Studied: English
0 Replies
Loading